[
{
"path": ".dockerignore",
"content": "# 版本控制\n.git\n.github\n\n# 构建产物（builder 阶段重新生成）\n.next\nnode_modules\n\n# 测试\ncoverage\ntests\nvitest.config.ts\ndocker-compose.test.yml\n\n# 运行时数据\nlogs\ndata\ncertificates\nbackups\nuploads\n\n# IDE 和 AI 工具\n.vscode\n.cursor\n.claude\n.gemini\n.agent\n.shared\n.artifacts\n\n# 数据库文件\n*.db\n*.db-journal\nprisma/data\n\n# 临时文件\n.DS_Store\n*.tsbuildinfo\n.tmp-old-snapshot-*\n\n# 环境变量\n.env\n.env.local\n.env.*.local\n.env.test\n\n# 文档和杂项\n*.md\n*.py\ndebug-request.json\nAGENTS.md\ndocs\nagent\n"
},
{
"path": ".eslintrc.json",
"content": "{\n \"extends\": \"next/core-web-vitals\",\n \"rules\": {\n \"@typescript-eslint/no-explicit-any\": \"off\",\n \"@typescript-eslint/no-unused-vars\": \"warn\",\n \"@next/next/no-img-element\": \"warn\",\n \"react-hooks/exhaustive-deps\": \"warn\"\n }\n}\n\n"
},
{
"path": ".gitignore",
"content": "# See https://help.github.com/articles/ignoring-files/ for more about ignoring files.\n\n# dependencies\n/node_modules\n/.pnp\n.pnp.*\n.yarn/*\n!.yarn/patches\n!.yarn/plugins\n!.yarn/releases\n!.yarn/versions\n\n# testing\n/coverage\n\n# next.js\n/.next/\n/out/\n\n# production\n/build\n\n# misc\n.DS_Store\n*.pem\n\n# debug\nnpm-debug.log*\nyarn-debug.log*\nyarn-error.log*\n.pnpm-debug.log*\n\n# vercel\n.vercel\n\n# IDE and AI tools\n.vscode/\n.idea/\n.claude/\n.cursor/\n.gemini/\n.artifacts/\n.agent/\n.shared/\n\n# typescript\n*.tsbuildinfo\nnext-env.d.ts\n\n/src/generated/prisma\n\n# logs\n/logs\n*.log\n\n# environment variables\n.env\n.env.local\n.env.test\n.env.*.local\ndocker-logs/\n\n# database\n*.db\n*.db-journal\nprisma/data/\n\n# uploads (user data)\nuploads/\n/data/\n\ncertificates\n\n# local temporary snapshots for old-version verification\n.tmp-old-snapshot-*/\n\n# temporary files\ntmp/\n\n# internal AI agent config (not for public)\nAGENTS.md\nagent/\n\n# internal docs (not for public)\ndocs/\n\n# GitHub CI workflows (internal)\n# .github/ — 只保留 workflows，忽略其他\n.github/*\n!.github/workflows/"
},
{
"path": ".husky/pre-commit",
"content": "#!/usr/bin/env sh\nset -e\n\nnpm run verify:commit\n"
},
{
"path": ".husky/pre-push",
"content": "#!/usr/bin/env sh\nset -e\n\nnpm run verify:push\n"
},
{
"path": ".nvmrc",
"content": "22.14.0\n"
},
{
"path": ".tmp_check_task.ts",
"content": "import { prisma } from '@/lib/prisma'\n\nconst id = 'a3cbc6d3-8720-4584-addd-e2bc4ace7759'\n\nasync function main() {\n const t = await prisma.task.findUnique({\n where: { id },\n select: {\n id: true,\n type: true,\n userId: true,\n projectId: true,\n payload: true,\n errorMessage: true,\n createdAt: true,\n },\n })\n console.log(JSON.stringify(t, null, 2))\n\n if (!t) return\n\n const pref = await prisma.userPreference.findUnique({\n where: { userId: t.userId },\n select: {\n analysisModel: true,\n customProviders: true,\n customModels: true,\n },\n })\n console.log('userPreference', JSON.stringify(pref, null, 2))\n}\n\nmain()\n .catch((err) => {\n console.error(err)\n process.exit(1)\n })\n .finally(async () => {\n await prisma.$disconnect()\n })\n"
},
{
"path": "CHANGELOG.md",
"content": "# Changelog / 更新日志\n\nAll notable changes to this project will be documented in this file.\n\n---\n\n## [v0.2] - 2026-02-28\n\n### ✨ 新功能\n- 增加 OpenAI 兼容图片、视频格式支持\n\n### 🐛 修复\n- 修复默认模型配置后项目模型需要二次选择的问题\n- 修复部分情况 resolution 无法读取的问题\n- 修复模型链路为 LangGraph\n- 修复默认参数无选择问题\n- 修复关闭计费依然触发计费问题\n- 修复 openai-compatible 被误判为原生 OpenAI 推理问题\n- 修复 JSON 解析失败问题\n\n### ⚙️ 优化\n- 修改为默认计费 off\n- 增强提示词 JSON 格式限制\n\n---\n\n## [v0.2.1] - 2026-02-28\n\n### 🐛 修复\n- 修复 AI 生成内容语言不跟随网站语言设置的问题\n- 修复前端 API 请求未携带 Accept-Language header 导致 locale 回退到浏览器默认语言\n---\n\n## [v0.1] - 2026-02-27\n\n### 🎉 首次发布\n- 项目初始开源版本\n"
},
{
"path": "Dockerfile",
"content": "# ==================== Stage 1: Dependencies ====================\nFROM node:20-alpine AS deps\nWORKDIR /app\n\nCOPY package.json package-lock.json ./\nCOPY prisma ./prisma\nRUN npm ci\n\n# ==================== Stage 2: Build ====================\nFROM node:20-alpine AS builder\nWORKDIR /app\n\nCOPY --from=deps /app/node_modules ./node_modules\nCOPY . .\n\n# Prisma generate + Next.js build\nRUN npm run build\n\n# ==================== Stage 3: Production ====================\nFROM node:20-alpine AS runner\nWORKDIR /app\n\nENV NODE_ENV=production\n\n# Install tini for proper signal handling\nRUN apk add --no-cache tini\n\n# node_modules（含 devDeps，因为 npm run start 需要 concurrently + tsx）\nCOPY --from=builder /app/node_modules ./node_modules\nCOPY --from=builder /app/package.json ./package.json\n\n# Next.js 构建产物\nCOPY --from=builder /app/.next ./.next\nCOPY --from=builder /app/public ./public\n\n# Prisma schema（db push 需要）\nCOPY --from=builder /app/prisma ./prisma\n\n# Worker 和 Watchdog 源码（tsx 运行 TypeScript）\nCOPY --from=builder /app/src ./src\nCOPY --from=builder /app/scripts ./scripts\nCOPY --from=builder /app/lib ./lib\n\n# 定价和配置标准\nCOPY --from=builder /app/standards ./standards\n\n# 国际化 + 配置文件\nCOPY --from=builder /app/messages ./messages\nCOPY --from=builder /app/tsconfig.json ./tsconfig.json\nCOPY --from=builder /app/next.config.ts ./next.config.ts\nCOPY --from=builder /app/middleware.ts ./middleware.ts\nCOPY --from=builder /app/postcss.config.mjs ./postcss.config.mjs\n\n# 运行日志目录 + 空 .env（tsx --env-file=.env 需要文件存在，实际 env 由 docker-compose 注入）\nRUN mkdir -p /app/logs && touch /app/.env\n\nEXPOSE 3000 3010\n\nENTRYPOINT [\"/sbin/tini\", \"--\"]\nCMD [\"npm\", \"run\", \"start\"]\n"
},
{
"path": "README.md",
"content": "\n \n \n
\n\n\n
\n\nwaoowaoo AI 影视 Studio \n\n\n 一款基于 AI 技术的短剧/漫画视频制作工具，支持从小说文本自动生成分镜、角色、场景，并制作成完整视频。\n
\n\n\n English · 加入内测候补 · 反馈问题 \n
\n\n> [!IMPORTANT]\n> ⚠️ **测试版声明**：本项目目前处于测试初期阶段，由于暂时只有我一个人开发，存在部分 bug 和不完善之处。我们正在快速迭代更新中，**欢迎进群反馈问题和需求，及时关注项目更新！目前更新会非常频繁，后续会增加大量新功能以及优化效果，我们的目标是成为行业最强AI工具！**\n\n\n
\n\nwaoowaoo AI Video Studio \n\n\n An AI-powered tool for creating short drama / comic videos — automatically generates storyboards, characters, and scenes from novel text, then assembles them into complete videos.\n
\n\n\n 中文文档 · Join Waitlist · Report Bug \n
\n\n> [!IMPORTANT]\n> **Beta Notice**: This project is currently in its early beta stage. As it is currently a solo-developed project, some bugs and imperfections are to be expected. We are iterating rapidly — please stay tuned for frequent updates! We are committed to rolling out a massive roadmap of new features and optimizations, with the ultimate goal of becoming the top-tier solution in the industry. Your feedback and feature requests are highly welcome!\n\n---\n\n## ✨ Features\n\n- 🎬 **AI Script Analysis** — Parse novels, extract characters, scenes & plot automatically\n- 🎨 **Character & Scene Generation** — Consistent AI-generated character and scene images\n- 📽️ **Storyboard Video** — Auto-generate shots and compose into complete videos\n- 🎙️ **AI Voiceover** — Multi-character voice synthesis\n- 🌐 **Bilingual UI** — Chinese / English, switch in the top-right corner\n\n---\n\n## 🚀 Quick Start\n\n**Prerequisites**: Install [Docker Desktop](https://docs.docker.com/get-docker/)\n\n### Method 1: Pull Pre-built Image (Easiest)\n\nNo need to clone the repository. Just download and run:\n\n```bash\n# Download docker-compose.yml\ncurl -O https://raw.githubusercontent.com/saturndec/waoowaoo/main/docker-compose.yml\n\n# Start all services\ndocker compose up -d\n```\n\n> ⚠️ This is a beta version. Database is not compatible between versions. To upgrade, clear old data first:\n\n```bash\ndocker compose down -v\ndocker rmi ghcr.io/saturndec/waoowaoo:latest\ncurl -O https://raw.githubusercontent.com/saturndec/waoowaoo/main/docker-compose.yml\ndocker compose up -d\n```\n\n> After starting, please **clear your browser cache** and log in again to avoid issues caused by stale cache.\n\n### Method 2: Clone & Docker Build (Full Control)\n\n```bash\ngit clone https://github.com/saturndec/waoowaoo.git\ncd waoowaoo\ndocker compose up -d\n```\n\nTo update:\n```bash\ngit pull\ndocker compose down && docker compose up -d --build\n```\n\n### Method 3: Local Development (For Developers)\n\n```bash\ngit clone https://github.com/saturndec/waoowaoo.git\ncd waoowaoo\n\n# Copy environment config (must be done before npm install)\ncp .env.example .env\n# ⚠️ Edit .env to fill in your AI API Keys (NEXTAUTH_URL defaults to http://localhost:3000, no change needed)\n\nnpm install\n\n# Start infrastructure only\ndocker compose up mysql redis minio -d\n\n# Run database migration\nnpx prisma db push\n\n# Start development server\nnpm run dev\n```\n\n---\n\nVisit [http://localhost:13000](http://localhost:13000) (Method 1 & 2) or [http://localhost:3000](http://localhost:3000) (Method 3) to get started!\n\n> The database is initialized automatically on first launch — no extra configuration needed.\n\n> [!TIP]\n> **If you experience lag**: HTTP mode may limit browser connections. Install [Caddy](https://caddyserver.com/docs/install) for HTTPS:\n> ```bash\n> caddy run --config Caddyfile\n> ```\n> Then visit [https://localhost:1443](https://localhost:1443)\n\n---\n\n## 🔧 API Configuration\n\nAfter launching, go to **Settings** to configure your AI service API keys. A built-in guide is provided.\n\n> 💡 **Note**: Currently only official provider APIs are recommended. Third-party compatible formats (OpenAI Compatible) are not yet fully supported and will be improved in future releases.\n\n---\n\n## 📦 Tech Stack\n\n- **Framework**: Next.js 15 + React 19\n- **Database**: MySQL + Prisma ORM\n- **Queue**: Redis + BullMQ\n- **Styling**: Tailwind CSS v4\n- **Auth**: NextAuth.js\n\n---\n\n## 📦 Preview\n\n\n\n\n\n\n---\n\n## 🤝 Contributing\n\nThis project is maintained by the core team. You're welcome to contribute by:\n\n- 🐛 Filing [Issues](https://github.com/saturndec/waoowaoo/issues) — report bugs\n- 💡 Filing [Issues](https://github.com/saturndec/waoowaoo/issues) — propose features\n- 🔧 Submitting Pull Requests as references — we review every PR carefully for ideas, but the team implements fixes internally rather than merging external PRs directly\n\n---\n\n**Made with ❤️ by waoowaoo team**\n\n## Star History\n\n[](https://www.star-history.com/#saturndec/waoowaoo&type=date&legend=top-left)\n"
},
{
"path": "caddyfile",
"content": "# HTTPS 反向代理（在主机上运行，非 Docker 内）\n# 启动方式: caddy run --config Caddyfile\n#\n# 用法:\n# 1. docker compose up -d (启动 App + MySQL + Redis)\n# 2. caddy run --config Caddyfile (启动 HTTPS 代理)\n# 3. 打开 https://localhost:4443 或 https://your-ip:4443\n#\n# 修改下方 IP 为你的局域网 IP（ifconfig en0 查看）\n# 例如: localhost:1443, https://192.168.x.x:1443 {\n\nlocalhost:1443 {\n handle /admin/queues* {\n reverse_proxy localhost:13010\n }\n handle {\n reverse_proxy localhost:13000\n }\n tls internal\n}"
},
{
"path": "debug-request.json",
"content": "{\n \"model\": \"doubao-seedream-4-0-250828\",\n \"prompt\": \"Lily and Olivia in 医院病房_日夜, medium shot, dramatic lighting, American comic style\",\n \"sequential_image_generation\": \"disabled\",\n \"response_format\": \"url\",\n \"size\": \"1080x1920\",\n \"stream\": false,\n \"watermark\": false\n}"
},
{
"path": "docker-compose.test.yml",
"content": "services:\n mysql:\n image: mysql:8.0\n container_name: waoowaoo-test-mysql\n environment:\n MYSQL_ROOT_PASSWORD: root\n MYSQL_DATABASE: waoowaoo_test\n MYSQL_ROOT_HOST: \"%\"\n ports:\n - \"3307:3306\"\n command:\n - \"--default-authentication-plugin=mysql_native_password\"\n - \"--sql_mode=STRICT_TRANS_TABLES,NO_ENGINE_SUBSTITUTION\"\n healthcheck:\n test: [\"CMD-SHELL\", \"mysqladmin ping -h 127.0.0.1 -uroot -proot\"]\n interval: 5s\n timeout: 5s\n retries: 30\n start_period: 15s\n\n redis:\n image: redis:7-alpine\n container_name: waoowaoo-test-redis\n ports:\n - \"6380:6379\"\n command: [\"redis-server\", \"--appendonly\", \"no\"]\n healthcheck:\n test: [\"CMD\", \"redis-cli\", \"ping\"]\n interval: 5s\n timeout: 5s\n retries: 30\n start_period: 5s\n"
},
{
"path": "docker-compose.yml",
"content": "services:\n # ==================== MySQL ====================\n mysql:\n image: mysql:8.0\n container_name: waoowaoo-mysql\n restart: unless-stopped\n environment:\n MYSQL_ROOT_PASSWORD: waoowaoo123\n MYSQL_DATABASE: waoowaoo\n MYSQL_ROOT_HOST: \"%\"\n ports:\n - \"13306:3306\"\n volumes:\n - mysql_data:/var/lib/mysql\n command:\n - \"--default-authentication-plugin=mysql_native_password\"\n - \"--sql_mode=STRICT_TRANS_TABLES,NO_ENGINE_SUBSTITUTION\"\n healthcheck:\n test: [\"CMD-SHELL\", \"mysqladmin ping -h 127.0.0.1 -uroot -pwaoowaoo123\"]\n interval: 5s\n timeout: 5s\n retries: 30\n start_period: 15s\n\n # ==================== Redis ====================\n redis:\n image: redis:7-alpine\n container_name: waoowaoo-redis\n restart: unless-stopped\n ports:\n - \"16379:6379\"\n volumes:\n - redis_data:/data\n command: [\"redis-server\", \"--appendonly\", \"yes\"]\n healthcheck:\n test: [\"CMD\", \"redis-cli\", \"ping\"]\n interval: 5s\n timeout: 5s\n retries: 30\n start_period: 5s\n\n # ==================== MinIO ====================\n minio:\n image: minio/minio:RELEASE.2025-02-28T09-55-16Z\n container_name: waoowaoo-minio\n restart: unless-stopped\n environment:\n MINIO_ROOT_USER: minioadmin\n MINIO_ROOT_PASSWORD: minioadmin\n command: server /data --console-address \":9001\"\n ports:\n - \"19000:9000\"\n - \"19001:9001\"\n volumes:\n - minio_data:/data\n healthcheck:\n test: [\"CMD\", \"curl\", \"-f\", \"http://127.0.0.1:9000/minio/health/live\"]\n interval: 5s\n timeout: 5s\n retries: 30\n start_period: 10s\n\n # ==================== App (Next.js + Workers) ====================\n app:\n image: ghcr.io/saturndec/waoowaoo:latest\n container_name: waoowaoo-app\n restart: unless-stopped\n environment:\n # 数据库（指向容器内部 MySQL，用服务名 mysql 而非 localhost）\n DATABASE_URL: \"mysql://root:waoowaoo123@mysql:3306/waoowaoo\"\n # Redis（指向容器内部 Redis，用服务名 redis）\n REDIS_HOST: redis\n REDIS_PORT: \"6379\"\n REDIS_USERNAME: \"\"\n REDIS_PASSWORD: \"\"\n REDIS_TLS: \"\"\n # 存储：默认 MinIO（S3 兼容）\n STORAGE_TYPE: minio\n MINIO_ENDPOINT: \"http://minio:9000\"\n MINIO_REGION: \"us-east-1\"\n MINIO_BUCKET: \"waoowaoo\"\n MINIO_ACCESS_KEY: \"minioadmin\"\n MINIO_SECRET_KEY: \"minioadmin\"\n MINIO_FORCE_PATH_STYLE: \"true\"\n # 外部访问地址（浏览器实际访问）\n NEXTAUTH_URL: \"http://localhost:13000\"\n # 容器内自调用地址（服务端 fetch 自己的 API / 文件）\n INTERNAL_APP_URL: \"http://127.0.0.1:3000\"\n NEXTAUTH_SECRET: \"waoowaoo-default-secret-2026\"\n # 内部密钥\n CRON_SECRET: \"waoowaoo-docker-cron-secret\"\n INTERNAL_TASK_TOKEN: \"waoowaoo-docker-task-token\"\n API_ENCRYPTION_KEY: \"waoowaoo-opensource-fixed-key-2026\"\n # Worker 配置\n WATCHDOG_INTERVAL_MS: \"30000\"\n TASK_HEARTBEAT_TIMEOUT_MS: \"90000\"\n QUEUE_CONCURRENCY_IMAGE: \"50\"\n QUEUE_CONCURRENCY_VIDEO: \"50\"\n QUEUE_CONCURRENCY_VOICE: \"20\"\n QUEUE_CONCURRENCY_TEXT: \"50\"\n # Bull Board\n BULL_BOARD_HOST: \"0.0.0.0\"\n BULL_BOARD_PORT: \"3010\"\n BULL_BOARD_BASE_PATH: \"/admin/queues\"\n BULL_BOARD_USER: \"\"\n BULL_BOARD_PASSWORD: \"\"\n # 日志\n LOG_UNIFIED_ENABLED: \"true\"\n LOG_LEVEL: \"INFO\"\n LOG_FORMAT: \"json\"\n LOG_DEBUG_ENABLED: \"false\"\n LOG_AUDIT_ENABLED: \"true\"\n LOG_SERVICE: \"waoowaoo\"\n LOG_REDACT_KEYS: \"password,token,apiKey,apikey,authorization,cookie,secret,access_token,refresh_token\"\n # 计费\n BILLING_MODE: \"OFF\"\n # 流式\n LLM_STREAM_EPHEMERAL_ENABLED: \"true\"\n ports:\n - \"13000:3000\"\n - \"13010:3010\"\n volumes:\n - ./data:/app/data\n - ./docker-logs:/app/logs\n depends_on:\n mysql:\n condition: service_healthy\n redis:\n condition: service_healthy\n minio:\n condition: service_healthy\n command: >\n sh -c \"\n npx prisma db push --skip-generate &&\n (sleep 5 && echo '' &&\n echo '╔══════════════════════════════════════════════════╗' &&\n echo '║ waoowaoo is ready! ║' &&\n echo '║ ║' &&\n echo '║ HTTP: http://localhost:13000 ║' &&\n echo '║ ║' &&\n echo '║ For HTTPS, run Caddy on host: ║' &&\n echo '║ caddy run --config Caddyfile ║' &&\n echo '║ Then open: https://localhost:1443 ║' &&\n echo '╚══════════════════════════════════════════════════╝' &&\n echo '') &\n npm run start\n \"\n\nvolumes:\n mysql_data:\n redis_data:\n minio_data:\n"
},
{
"path": "eslint.config.mjs",
"content": "import { dirname } from \"path\";\nimport { fileURLToPath } from \"url\";\nimport { FlatCompat } from \"@eslint/eslintrc\";\n\nconst __filename = fileURLToPath(import.meta.url);\nconst __dirname = dirname(__filename);\n\nconst compat = new FlatCompat({\n baseDirectory: __dirname,\n});\n\nconst eslintConfig = [\n ...compat.extends(\"next/core-web-vitals\", \"next/typescript\"),\n {\n ignores: [\n \"node_modules/**\",\n \".agent/**\",\n \".next/**\",\n \"out/**\",\n \"build/**\",\n \"coverage/**\",\n \"next-env.d.ts\",\n ],\n },\n {\n files: [\"src/**/*.{ts,tsx}\"],\n ignores: [\"src/components/ui/icons/**\"],\n rules: {\n \"no-restricted-imports\": [\n \"error\",\n {\n paths: [\n {\n name: \"lucide-react\",\n message: \"Import icons through '@/components/ui/icons' only.\",\n },\n ],\n },\n ],\n \"no-restricted-syntax\": [\n \"error\",\n {\n selector: \"JSXOpeningElement[name.name='svg']\",\n message:\n \"Use AppIcon or icons module components instead of inline .\",\n },\n ],\n },\n },\n];\n\nexport default eslintConfig;\n"
},
{
"path": "extract_chinese.py",
"content": "#!/usr/bin/env python3\n\"\"\"\n提取React/TypeScript代码中的硬编码中文字符串\n\"\"\"\nimport re\nimport os\nfrom pathlib import Path\nimport json\n\ndef extract_chinese_strings(file_path):\n \"\"\"提取文件中的中文字符串\"\"\"\n try:\n with open(file_path, 'r', encoding='utf-8') as f:\n content = f.read()\n except:\n return []\n \n results = []\n \n # 匹配JSX/TSX中的中文字符串\n # 1. {' 中文 '} 或 {\"中文\"}\n pattern1 = r'\\{\\s*[\\'\"]([^\\'\"\\{\\}]*[\\u4e00-\\u9fff]+[^\\'\"\\{\\}]*)[\\'\\\"]\\s*\\}'\n # 2. >中文< \n pattern2 = r'\\>([^<\\>]*[\\u4e00-\\u9fff]+[^<\\>]*)\\<'\n # 3. placeholder=\"中文\" 等属性\n pattern3 = r'(?:placeholder|title|alt|value|defaultValue|confirmText|cancelText|message)\\s*=\\s*[\\'\"]([^\\'\\\"]*[\\u4e00-\\u9fff]+[^\\'\\\"]*)[\\'\"]'\n # 4. 字符串默认值 = '中文'\n pattern4 = r'=\\s*[\\'\"]([^\\'\\\"]*[\\u4e00-\\u9fff]+[^\\'\\\"]*)[\\'\"]'\n \n for pattern in [pattern1, pattern2, pattern3, pattern4]:\n matches = re.finditer(pattern, content)\n for match in matches:\n chinese_text = match.group(1).strip()\n if chinese_text and len(chinese_text) > 0:\n # 跳过注释\n line_num = content[:match.start()].count('\\n') + 1\n line = content.split('\\n')[line_num - 1]\n if '//' in line and line.index('//') < line.find(chinese_text):\n continue\n results.append({\n 'text': chinese_text,\n 'line': line_num,\n 'category': 'unknown'\n })\n \n # 去重\n seen = set()\n unique_results = []\n for r in results:\n key = f\"{r['text']}_{r['line']}\"\n if key not in seen:\n seen.add(key)\n unique_results.append(r)\n \n return unique_results\n\ndef scan_directory(base_path,exclude_patterns=['test-ui']):\n \"\"\"扫描目录中的所有TSX/TS文件\"\"\"\n all_findings = {}\n \n for root, dirs, files in os.walk(base_path):\n # 排除特定目录\n dirs[:] = [d for d in dirs if d not in exclude_patterns and not d.startswith('.')]\n \n for file in files:\n if file.endswith(('.tsx', '.ts')):\n file_path = os.path.join(root, file)\n relative_path = os.path.relpath(file_path, base_path)\n \n findings = extract_chinese_strings(file_path)\n if findings:\n all_findings[relative_path] = findings\n \n return all_findings\n\nif __name__ == '__main__':\n base_dir = 'src'\n results = scan_directory(base_dir)\n \n # 输出结果\n total = 0\n for file_path, findings in sorted(results.items()):\n if findings:\n print(f\"\\n## {file_path} ({len(findings)} strings)\")\n for finding in findings[:10]: # 只显示前10个\n print(f\" Line {finding['line']}: {finding['text'][:60]}\")\n total += len(findings)\n if len(findings) > 10:\n print(f\" ... and {len(findings) - 10} more\")\n \n print(f\"\\n\\n总计: {len(results)} 个文件, {total} 处硬编码中文\")\n"
},
{
"path": "lib/prompts/character-reference/character_image_to_description.en.txt",
"content": "# Character Image To Description Prompt\n\nAnalyze the provided character image and write one detailed English visual description for image generation.\n\n## Required content\nInclude all of the following:\n1. Gender and approximate age range\n2. Hair style and hair color\n3. Face shape and facial features\n4. Body build and silhouette\n5. Clothing style and clothing details\n6. Accessories and signature details\n7. Overall style keywords\n\n## Missing-part completion\nIf the image only shows part of the body, infer the missing parts consistently.\n- Missing lower body: infer matching pants/skirt style\n- Missing shoes: infer shoes that fit the outfit\n- Missing accessory details: infer a few reasonable accessories\n\n## Forbidden content\nDo not mention:\n- Skin color\n- Eye color\n- Facial expression\n- Pose or action\n- Background\n\n## Output format\nReturn one plain English paragraph only (about 120-220 words).\nDo not return markdown, bullets, titles, or JSON.\n"
},
{
"path": "lib/prompts/character-reference/character_image_to_description.zh.txt",
"content": "# 图片反推角色描述提示词\n\n请分析这张角色图片，生成一段详细的角色外貌描述（用于 AI 图片生成）。\n\n## 输出要求\n\n生成一段完整的角色视觉描述，包含以下要素：\n\n1. 性别和年龄段（如：约二十五岁的男性）\n2. 发型发色（如：黑色短发、微卷的棕色长发）\n3. 脸型五官特征（如：剑眉星目、高鼻梁、薄唇）\n4. 体态身材（如：身形修长、体格健壮）\n5. 服装风格（如：深蓝色西装、白色衬衫、皮质腰带）\n6. 配饰特征（如：左手戴银色手表、胸前别金色胸针）\n7. 整体气质关键词（如：精英气质、禁欲系、高冷、温柔暖男）\n\n## 缺失内容补齐规则\n\n如果参考图只展示了部分身体（如上半身、头像），请根据已有信息合理推断并补全：\n- **缺少下半身**：根据上衣风格推断裤装/裙装类型（如西装上衣配深蓝色西裤、休闲上衣配牛仔裤）\n- **缺少鞋子**：根据整体穿搭风格推断鞋款（如正装配皮鞋、休闲装配运动鞋或帆布鞋）\n- **缺少配饰细节**：根据角色气质合理添加配饰（如商务风配手表、休闲风配手环）\n\n## 禁止描写\n\n- 皮肤颜色\n- 眼睛颜色\n- 表情\n- 动作\n- 背景\n- 姿势\n\n## 输出格式\n\n一段连贯的描述文字，约200-300字，直接可用于图片生成提示词。\n只返回描述文字，不要有任何标题、序号或其他格式。\n"
},
{
"path": "lib/prompts/character-reference/character_reference_to_sheet.en.txt",
"content": "# Reference Image To Character Sheet Prompt (img2img)\n\nUse the provided reference image to extract face traits, hairstyle, body shape, and outfit structure.\n\n## Style priority\nFollow the user-selected style instruction as the highest-priority rule.\nUse reference images only to preserve character identity traits (face, hairstyle, body shape, outfit structure), and do not let reference style override the requested style.\nOnly if no explicit style instruction is provided, you may preserve the original visual style.\n\n## Generation rules\n1. Ignore the original image color cast and lighting defects\n2. Use clean, soft studio lighting\n3. Keep natural, aesthetically correct body proportions\n4. Do not copy blur, noise, compression artifacts, or defects\n5. Output must be clear, sharp, and production-quality\n6. Character expression should be neutral and calm, looking at camera\n\n## Missing-part completion\nIf only half body or partial body is visible, infer and complete hidden parts consistently:\n- Infer matching lower-body clothing\n- Infer suitable footwear\n- Infer missing hands/arms in a natural way\n"
},
{
"path": "lib/prompts/character-reference/character_reference_to_sheet.zh.txt",
"content": "# 参考图转角色设定图提示词（图生图模式）\n\n基于提供的参考图片，提取角色的面部五官特征、发型、体型和服装款式作为参考。\n\n## 画风优先级规则\n\n画风由用户选择的风格指令决定，严格遵循风格指令进行生成。\n参考图仅用于保持角色身份特征（五官、发型、体型、服装结构），不能覆盖用户指定画风。\n仅在未提供风格指令时，才可参考原图画风。\n\n## 生成规则\n\n1. 忽略原图的具体色调和光线\n2. 使用自然柔和的摄影棚灯光\n3. 绘制正常美观的人体比例\n4. 不要复制原图的画质、模糊、噪点或瑕疵\n5. 生成的图像必须清晰锐利、细节丰富、专业品质\n6. 角色表情应为自然平静的中性表情，目光正视镜头\n\n## 缺失部位自动补齐\n\n如果参考图是半身或部分身体，请根据服装风格和人物特征合理补全未露出的部位：\n- **缺少下半身**：根据上衣风格推断并绘制匹配的裤装/裙装\n- **缺少脚部**：根据整体穿搭风格添加合适的鞋款\n- **缺少手部/手臂**：根据姿态合理补全\n- 保持整体风格一致，确保补全的部分与可见部分协调统一\n"
},
{
"path": "lib/prompts/novel-promotion/agent_acting_direction.en.txt",
"content": "You are an experienced Acting Director.\nYour task is to generate acting notes for each character in each panel.\n\nCore input:\n- Total panel count: {panel_count}\n- Panels JSON:\n{panels_json}\n- Character info:\n{characters_info}\n\nRequirements:\n1. Treat each panel independently. The same character can have different emotional states across panels.\n2. Adapt performance style to panel.scene_type (daily / emotion / action / epic / suspense).\n3. For each character, write one concise visual acting instruction including:\n - emotional state (visible, not abstract)\n - facial expression details\n - body language / posture\n - micro action and gaze direction\n4. Use only observable descriptions. Avoid abstract words like \"sad\" without visual evidence.\n\nOutput format (JSON array only):\n[\n {\n \"panel_number\": 1,\n \"characters\": [\n {\n \"name\": \"Character Name\",\n \"acting\": \"One-sentence visual acting direction\"\n }\n ]\n }\n]\n\nStrict constraints:\n1. Return JSON only, no markdown.\n2. Array length must equal {panel_count}.\n3. Each character object must contain only \"name\" and \"acting\".\n4. Keep character names exactly consistent with panel input.\n5. ⚠️ JSON SAFETY: All quotation marks in text (\"\"''「」 etc.) MUST be converted to corner brackets「」in JSON string values. NEVER use raw ASCII double quotes \" inside string values.\n"
},
{
"path": "lib/prompts/novel-promotion/agent_acting_direction.zh.txt",
"content": "你是一位经验丰富的表演指导(Acting Director)。你的任务是为一组分镜中的**每个镜头**设计角色的表演细节。\n\n【核心职责】\n\n分析整组分镜后，为每个镜头中的角色用一句话描述完整的表演指令，包含：\n- 情绪状态与强度\n- 面部表情细节\n- 肢体语言与姿态\n- 微动作与视线\n\n【重要】每个镜头的表演必须是独立的！\n- 同一角色在不同镜头可能有不同情绪变化\n- 表演风格匹配 scene_type（日常/情感/动作/史诗/悬疑）\n\n【表演风格匹配 scene_type】\n\n**daily（日常）**：自然松弛，微表情为主，动作幅度小\n**emotion（情感）**：细腻层次，眼神戏份重，情绪渐进\n**action（动作）**：爆发力强，动作干脆，表情夸张\n**epic（史诗）**：庄重仪式感，姿态端正，动作缓慢有力\n**suspense（悬疑）**：紧绷警觉，肢体僵硬，眼神游移\n\n【表演描述词库】\n\n**表情**：眼眶泛红、眉头紧锁、嘴角上扬、目光闪躲、瞳孔收缩、嘴唇颤抖、咬紧牙关\n**肢体**：握紧拳头、身体前倾、双手交握、肩膀耸起、转身背对、后退一步\n**微动作**：轻轻眨眼、咽口水、深呼吸、手指轻颤、舔嘴唇、胸口起伏\n\n【⚠️ 禁止规则】\n\n1. 禁止抽象情绪词：悲伤、愤怒、紧张 → 改用可见表现\n2. 禁止身份称呼：母亲、父亲 → 改用角色名\n\n【输出格式】\n\n返回JSON数组，每个镜头一个对象，每个角色只有 name + acting 两个字段：\n\n[\n {\n \"panel_number\": 1,\n \"characters\": [\n {\n \"name\": \"李凤华\",\n \"acting\": \"嘴角微扬眼神柔和地看向景笙，身体微微前倾，双手自然垂放，轻轻眨眼\"\n },\n {\n \"name\": \"景笙\",\n \"acting\": \"面带微笑但眼神略显疏离，站姿笔直双手背后，轻轻点头\"\n }\n ]\n },\n {\n \"panel_number\": 2,\n \"characters\": [\n {\n \"name\": \"李凤华\",\n \"acting\": \"眉头轻皱嘴唇紧抿，双手交握在身前肩膀微耸，目光低垂看向地面，手指轻微交缠\"\n }\n ]\n },\n {\n \"panel_number\": 3,\n \"characters\": [\n {\n \"name\": \"李凤华\",\n \"acting\": \"眼眶泛红泪水打转嘴唇颤抖，身体微微发抖双手攥紧衣角，快速眨眼忍住泪水转头避开对方视线\"\n }\n ]\n }\n]\n\n【输入数据】\n\n分镜数据（共 {panel_count} 个镜头）：\n{panels_json}\n\n角色信息：\n{characters_info}\n\n【严格要求】\n\n1. 只返回JSON数组，不要有markdown代码块标记\n2. 数组长度必须等于输入的镜头数量（{panel_count}个）\n3. 每个角色只有 name 和 acting 两个字段\n4. acting 用一句话描述完整表演（表情+肢体+微动作+视线）\n5. 角色名必须与输入分镜中的 characters 完全一致\n6. 所有描述必须是可视化的，禁止抽象情绪词\n7. 根据 scene_type 调整表演风格\n8. 情绪弧线连贯，前后镜头有合理递进\n9. ⚠️ JSON安全：所有引号（\"\"''等）在 JSON 字符串值中必须统一替换为「」，严禁出现未转义的英文双引号 \"\n\n"
},
{
"path": "lib/prompts/novel-promotion/agent_character_profile.en.txt",
"content": "You are a casting and character-asset analyst.\nAnalyze the input text and produce structured character profiles for visual production.\n\nInput text:\n{input}\n\nExisting character library info:\n{characters_lib_info}\n\nGoals:\n1. Identify characters that should appear visually.\n2. Exclude pure background extras and abstract entities.\n3. Build profile fields needed for downstream visual generation.\n4. Capture naming/alias mapping, especially first-person references.\n\nExtraction rules:\n1. Include characters that speak, act, or significantly drive plot.\n2. Exclude one-off nameless background people unless visual identity is required.\n3. Resolve aliases/titles (e.g., \"my husband\", \"boss\", \"I\") to canonical names when possible.\n4. For first-person narrative, explicitly document who \"I\" maps to in introduction.\n\nField rules:\n- role_level: one of S/A/B/C/D based on narrative importance\n- costume_tier: 1-5 based on social identity (not role_level)\n- expected_appearances: always include at least one initial appearance\n- introduction: include identity, relationship mapping, and common address mapping\n\nOutput format (JSON only):\n{\n \"characters\": [\n {\n \"name\": \"Canonical Name\",\n \"aliases\": [\"alias 1\", \"alias 2\"],\n \"introduction\": \"Role, perspective mapping, relationships, and naming aliases\",\n \"gender\": \"male/female/other\",\n \"age_range\": \"young adult\",\n \"role_level\": \"S\",\n \"archetype\": \"character archetype\",\n \"personality_tags\": [\"tag1\", \"tag2\"],\n \"era_period\": \"modern/fantasy/historical/sci-fi\",\n \"social_class\": \"elite/middle/common\",\n \"occupation\": \"occupation or none\",\n \"costume_tier\": 3,\n \"suggested_colors\": [\"color1\", \"color2\"],\n \"primary_identifier\": \"signature visual marker\",\n \"visual_keywords\": [\"keyword1\", \"keyword2\"],\n \"expected_appearances\": [\n { \"id\": 1, \"change_reason\": \"initial appearance\" }\n ]\n }\n ],\n \"new_characters\": [\n {\n \"name\": \"Canonical Name\",\n \"aliases\": [\"alias 1\", \"alias 2\"],\n \"introduction\": \"Role, perspective mapping, relationships, and naming aliases\",\n \"gender\": \"male/female/other\",\n \"age_range\": \"young adult\",\n \"role_level\": \"S\",\n \"archetype\": \"character archetype\",\n \"personality_tags\": [\"tag1\", \"tag2\"],\n \"era_period\": \"modern/fantasy/historical/sci-fi\",\n \"social_class\": \"elite/middle/common\",\n \"occupation\": \"occupation or none\",\n \"costume_tier\": 3,\n \"suggested_colors\": [\"color1\", \"color2\"],\n \"primary_identifier\": \"signature visual marker\",\n \"visual_keywords\": [\"keyword1\", \"keyword2\"],\n \"expected_appearances\": [\n { \"id\": 1, \"change_reason\": \"initial appearance\" }\n ]\n }\n ],\n \"updated_characters\": [\n {\n \"name\": \"Existing Canonical Name\",\n \"updated_introduction\": \"Updated intro with newly discovered mapping\",\n \"updated_aliases\": [\"new alias 1\", \"new alias 2\"]\n }\n ]\n}\n\nStrict constraints:\n1. JSON only.\n2. If no valid character is found, return empty arrays.\n3. Keep all strings in English.\n4. ⚠️ JSON SAFETY: All quotation marks in text (\"\"''「」 etc.) MUST be converted to corner brackets「」in JSON string values. NEVER use raw ASCII double quotes \" inside string values.\n"
},
{
"path": "lib/prompts/novel-promotion/agent_character_profile.zh.txt",
"content": "你是专业的\"选角指导\"。请基于提供的文本（小说、剧本或混合格式），分析并输出所有需要制作形象的角色档案信息。\n\n【你的职责】\n- 识别需要在画面中出现的角色\n- 根据剧情发展和角色身份判断每个角色的重要性层级\n- 分析角色的性格和背景\n- 输出结构化的角色档案（供后续视觉生成使用）\n- ⚠️ 分析角色之间的关系、称呼映射，生成角色介绍（introduction）\n\n【筛选规则 - 精准提取模式】\n\n✅【必须提取的角色】：\n - 剧本人物行中列出的角色\n - 有台词且参与剧情互动的角色\n - 贯穿故事主线的核心人物\n - 对剧情有实际推动作用的配角\n - 在画面中需要出镜的角色\n\n❌【不提取的角色】：\n - 无名无特征的纯路人（如\"人群中的某人\"）\n - 仅被提及但从未出场的角色\n - 没有台词也没有互动的背景人物\n - 意境描述中的虚构存在（如\"命运\"、\"死神的化身\"）\n\n📋【判断标准】：\n 问自己：这个角色是否需要制作形象图？是否在画面中有实际出镜？\n 如果答案是否定的，则不提取。\n\n【角色介绍 introduction 规则 ⭐重要】\n\n每个角色必须有 introduction 字段，用于帮助后续 AI 正确识别角色。包含：\n\n1. **叙述视角映射**：\n - 如果是第一人称叙述，明确说明\"我\"对应此角色\n - 示例：\"本角色是故事主角，小说以第一人称'我'叙述\"\n\n2. **角色身份定位**：\n - 描述角色在故事中的身份（主角/配角/反派等）\n - 示例：\"女主角，公司秘书\"\n\n3. **角色关系**：\n - 与其他主要角色的关系\n - 示例：\"林墨的妻子，张三的女儿\"\n\n4. **称呼映射**：\n - 其他角色对此角色的常用称呼\n - 示例：\"被林墨称呼为'老婆'、'晴晴'，被张三称呼为'闺女'\"\n\n示例 introduction：\n\"故事主角，小说以第一人称'我'叙述，真名林墨。苏晴的丈夫，张三的女婿。被苏晴称呼为'老公'、'墨哥'，被下属称呼为'林总'。\"\n\n【角色重要性层级判断规则】\n\n⚠️ 重要：根据角色在剧情中的戏份和身份地位来判断，不是根据外表华丽程度！\n\nS级（绝对主角）：\n - 故事的核心视角人物，剧情围绕其展开\n - 第一人称叙述中的\"我\"通常是S级\n - 判断依据：戏份最重、出场最多、剧情主线与其紧密相关\n\nA级（核心配角）：\n - 与主角有大量互动的重要角色\n - 男二号、女二号、主要反派等\n - 判断依据：对主线剧情有重大影响、戏份仅次于主角\n\nB级（重要配角）：\n - 多次出场、有名有姓、推动某条支线剧情\n - 判断依据：有一定戏份、对剧情有贡献\n\nC级（次要角色）：\n - 偶尔出场、戏份较少但有具体形象\n - 判断依据：需要出镜但戏份不多\n\nD级（群众演员）：\n - 有短暂出镜需求的小角色\n - 判断依据：仅在个别场景出现\n\n【服装华丽度层级 costume_tier】\n\n⚠️ 服装华丽度由角色的社会身份和剧情设定决定，不是由重要性决定！\n - 主角可以是朴素穿着（如穷学生主角=tier 2）\n - 配角可以是华丽服装（如富家公子配角=tier 5）\n\n5级（皇室/顶奢级）：皇室成员、顶级富豪，服装极致华丽，有精美的刺绣、镶嵌或定制剪裁。\n4级（贵族/精英级）：贵族、企业家，服装精致考究，使用高档面料和精致细节。\n3级（专业/品质级）：中产阶级、专业人士，服装得体有品，剪裁讲究。\n2级（日常/普通级）：普通人、学生，服装简洁日常，款式普通但整洁。\n1级（朴素/统一级）：平民、底层劳动者，服装朴素统一，基础款式，功能性为主。\n\n【角色原型 archetype 参考词库】\n\n正派角色可以选择：霸道总裁、高冷学霸、温柔暖男、励志少年、贤惠女主、独立女强人、忠诚护卫等。\n\n反派角色可以选择：心机婊、白莲花、阴险反派、疯批美人、复仇者等。\n\n其他类型：傲娇公主、病娇、腹黑、毒舌、话痨、冷面热心、闷骚等。\n\n【性格标签 personality_tags 参考词库】\n\n气质类标签：高冷、温柔、阳光、忧郁、神秘、妩媚、清冷、热情\n\n性格类标签：腹黑、傲娇、毒舌、话痨、闷骚、直爽、圆滑、固执\n\n态度类标签：自信、自卑、孤僻、合群、叛逆、顺从\n\n【视觉关键词 visual_keywords 参考词库】\n\n风格类关键词：精英气质、街头潮流、学院风、复古优雅、运动活力、文艺气息、冷淡极简\n\n特征类关键词：病弱感、禁欲系、狼狗系、奶狗系、御姐范、萝莉感、大叔味\n\n【色彩建议规则】\n\n根据角色类型选择合适的色彩：\n\n正派主角适合白色、蓝色、金色或浅色系，传达正义和光明感。\n\n反派角色适合黑色、暗红、深紫或暗色系，营造神秘或压迫感。\n\n温柔角色适合米白、淡粉、浅绿等柔和色，体现温暖亲和。\n\n冷酷角色适合黑色、灰色、深蓝等冷色调，强调距离感。\n\n活泼角色适合橙色、黄色等亮色系，展现活力和热情。\n\n【辨识标志设计规则】\n\n为S级和A级角色设计一眼就能认出的标志性特征：\n\n面部标志：眼角泪痣、剑眉、刀疤、胎记等独特面部特征。\n\n发型标志：白发、挑染、独特发型、发带等醒目的头发特征。\n\n服装标志：永远穿红色、标志性围巾、招牌外套等固定的服装元素。\n\n配饰标志：家传戒指、从不摘下的项链、拐杖等标志性物品。\n\n【子形象筛选规则 - 识别视觉外观变化 ⭐重要】\n\n分析原文中角色是否有多个视觉形态，输出到 expected_appearances 字段。\n\n✅ 需要记录的子形象（视觉上可见的变化）：\n - 衣着变化：换装、更换正装/休闲装、穿戴盔甲等\n - 年龄变化：穿越、回忆场景中的年轻/年老状态\n - 特殊装扮：出浴（围浴巾）、冒充他人的装扮\n - 发型改变：剪头、编发、盘发、披发等持续性外观变化\n\n❌ 不需要记录的（非视觉或临时状态）：\n - 情绪/心理状态：生气、开心、难过、紧张\n - 健康状态：生病、发烧（除非有明显视觉特征如绷带）\n - 临时动作：跑步、跳跃、战斗姿势\n - 模糊描述：\"蒙上了一层阴影\"\"眼神变了\"等抽象描述\n - 临时特效/光影状态：散发光芒、身上发光、气场外放、浑身火焰、佛光环绕、金光闪闪等后期可添加的特效\n - 战斗技能释放：发功、运功、施法、放大招、释放法术等技能状态\n - 一次性瞬间状态：被打飞、摔倒、中招、受击等不持续的状态\n\n⚠️ 判断标准：\n - 如果一个状态无法通过换装来体现，就不需要记录\n - 如果一个状态是通过后期特效（如发光、粒子、光环、火焰等）来表现的，不需要记录\n - 如果一个状态只持续几秒而非整个场景，不需要记录\n - 只有持续性的、需要重新制作人物形象图的外观变化才需要记录\n\n📋 expected_appearances 格式：\n - 每个角色必须至少有一个 id=1 的\"初始形象\"\n - 如有换装/年龄变化等，添加 id=2, 3... 的子形象\n - change_reason 简要说明变化原因（如\"出浴状态\"、\"战斗装束\"、\"年老回忆\"）\n\n【已有资产库】\n\n⚠️ 重要：请仔细阅读已有角色的介绍，判断新发现的角色名是否与已有角色是同一人！\n\n{characters_lib_info}\n\n【输出格式 - 支持新增和更新】\n\n只返回JSON，禁止任何markdown标记或注释。\n\n输出包含两个数组：\n- new_characters: 新发现的角色\n- updated_characters: 需要更新介绍的已有角色（如发现了新的称呼、关系、或真名）\n\n{\n \"new_characters\": [\n {\n \"name\": \"角色名\",\n \"aliases\": [\"别名1\", \"别名2\"],\n \"introduction\": \"角色介绍：身份定位、叙述视角映射、与其他角色的关系、常用称呼\",\n \"gender\": \"男/女\",\n \"age_range\": \"约二十五岁\",\n \"role_level\": \"S/A/B/C/D\",\n \"archetype\": \"角色原型（如霸道总裁）\",\n \"personality_tags\": [\"高冷\", \"腹黑\"],\n \"era_period\": \"现代都市/古代唐朝/未来科幻\",\n \"social_class\": \"上层精英/中产/平民\",\n \"occupation\": \"企业家/学生/无\",\n \"costume_tier\": 5,\n \"suggested_colors\": [\"深蓝\", \"金色\"],\n \"primary_identifier\": \"眼角泪痣（仅S/A级需要）\",\n \"visual_keywords\": [\"精英气质\", \"禁欲系\"],\n \"expected_appearances\": [\n {\"id\": 1, \"change_reason\": \"初始形象\"},\n {\"id\": 2, \"change_reason\": \"换装/特殊状态的原因（如有）\"}\n ]\n }\n ],\n \"updated_characters\": [\n {\n \"name\": \"已有角色名（必须与资产库中的名字完全一致）\",\n \"updated_introduction\": \"更新后的角色介绍（补充新发现的关系、称呼、真名等）\",\n \"updated_aliases\": [\"新发现的别名1\", \"新发现的别名2\"]\n }\n ]\n}\n\n【更新规则】\n\n⚠️ 什么情况下应该更新已有角色（放入 updated_characters）：\n\n1. **发现真名**：之前只有\"我\"，现在发现\"我\"的真名是\"林墨\"\n → 更新 introduction 说明映射，添加 updated_aliases: [\"林墨\"]\n\n2. **发现新称呼**：之前不知道别人怎么称呼这个角色，现在发现有人叫他\"林总\"\n → 更新 introduction 添加称呼信息，添加 updated_aliases: [\"林总\"]\n\n3. **发现新关系**：之前不知道角色间的关系，现在发现苏晴是林墨的妻子\n → 更新双方的 introduction 添加关系信息\n\n4. **不要重复创建**：如果发现\"林墨\"其实就是已有的\"我\"，不要创建新角色，而是更新\"我\"的介绍和别名\n\n【严格要求】\n1. 只返回JSON，不得有其他文字\n2. role_level 必须是 S/A/B/C/D 之一\n3. costume_tier 必须是 1-5 的整数\n4. S/A 级角色必须有 primary_identifier\n5. personality_tags 至少2个，最多5个\n6. suggested_colors 2-3个颜色\n7. introduction 必填，描述角色身份、关系、称呼映射\n8. 如果发现已有角色的新信息，放入 updated_characters 而不是创建新角色\n9. updated_characters 中的 name 必须与已有资产库中的名字完全一致\n10. expected_appearances 必填，至少包含 id=1 的初始形象\n11. 只有持续性视觉变化才添加子形象，临时特效/情绪/动作不添加\n12. 输出必须是**严格合法的JSON**：字符串中不能出现原始换行/回车/制表符\n\n⚠️⚠️⚠️【JSON安全输出 - 最高优先级】⚠️⚠️⚠️\n- 原文中的所有引号（\"\"''「」『』等）在 JSON 字符串值中必须统一替换为日式方括号引号「」\n- ❌ 严禁在 JSON 字符串值中出现英文双引号 \" ！会破坏 JSON 结构！\n- ✅ 正确：\"introduction\":\"他被称为「弼马温」\"\n- ❌ 错误：\"introduction\":\"他被称为\"弼马温\"\" ← 内部裸引号破坏JSON\n- 如果字符串内确实需要英文双引号，必须用 \\\" 转义\n\n【原文内容】\n{input}\n"
},
{
"path": "lib/prompts/novel-promotion/agent_character_visual.en.txt",
"content": "You are a character visual designer.\nGenerate image-ready appearance descriptions from character profiles.\n\nCharacter profiles JSON:\n{character_profiles}\n\nRules:\n1. Keep identity consistency with profile fields.\n2. Convert personality/social identity into visual details (face, hair, outfit, accessories).\n3. Support both human and non-human characters.\n4. Respect era_period and costume_tier.\n5. If primary_identifier exists, include it clearly in each main description.\n6. Do not include expression, action, background, or plot narration.\n7. Do not include skin color, eye color, or lip color.\n\nAppearance strategy:\n- Initial appearance should be complete and self-contained.\n- Additional appearances should focus on visual changes indicated by change_reason.\n- Provide 3 alternative description lines per appearance.\n\nOutput format (JSON only):\n{\n \"characters\": [\n {\n \"name\": \"Character Name\",\n \"appearances\": [\n {\n \"id\": 0,\n \"change_reason\": \"initial appearance\",\n \"descriptions\": [\n \"description variant 1\",\n \"description variant 2\",\n \"description variant 3\"\n ]\n }\n ]\n }\n ]\n}\n\nStrict constraints:\n1. JSON only.\n2. Keep names exactly aligned with input profiles.\n3. Each descriptions array must contain at least 3 valid English strings.\n4. ⚠️ JSON SAFETY: All quotation marks in text (\"\"''「」 etc.) MUST be converted to corner brackets「」in JSON string values. NEVER use raw ASCII double quotes \" inside string values.\n"
},
{
"path": "lib/prompts/novel-promotion/agent_character_visual.zh.txt",
"content": "你是专业的\"角色视觉设计师\"。根据角色档案信息，生成详细的人物外貌描述（用于AI图片生成）。\n\n【你的职责】\n- 根据角色档案生成对应的外貌描述\n- 确保核心角色有明显的视觉辨识度\n- 体现角色性格和身份的视觉特征\n- 服装华丽度由角色身份决定，与重要性无关\n\n【角色类型灵活处理规则】\n\n⚠️ 角色不一定是人类！请根据原文判断角色的实际形态：\n\n**人类角色**：按照下方的面部、发型、体态、服装规范描述\n\n**非人类角色**（动物、神话生物、知名形象等）：\n- 描述开头必须以角色名/物种名开始\n- 根据角色实际形态自由描述外观特征，不受人类模板限制\n- 保持角色的核心辨识特征\n\n示例：\n- 孙悟空 → \"孙悟空，身穿虎皮裙，头戴紧箍咒金环，手持如意金箍棒，毛发金黄蓬松，尖耳竖立，眼神机灵狡黠...\"\n- 蜗牛 → \"蜗牛，背负螺旋形褐色硬壳，壳面有细密纹路，两只细长触角顶端有圆形眼点，身体柔软半透明...\"\n- 龙 → \"东方神龙，鳞片金红交错闪烁，龙须飘逸，鹿角威严分叉，蛇身盘旋腾空，四爪锋利如钩...\"\n- 拟人化动物 → \"狐狸精，保留尖耳毛尾的狐狸特征，身着红色丝绸长裙，九条白色蓬松尾巴在身后舒展...\"\n\n【视觉层级规范】\n\n⚠️ 核心原则：服装华丽度由角色的社会身份和剧情设定决定，不是由重要性等级决定！\n\nS级角色：\n - 描述长度180-220字\n - 必须有极高的视觉辨识度和\"主角气质\"\n - 服装风格由角色身份决定（穷学生可以穿简单校服，但五官气质必须出众）\n\nA级角色：\n - 描述长度150-180字\n - 有明显的个人特色和记忆点\n - 服装风格由角色身份决定\n\nB级角色：\n - 描述长度120-150字\n - 有基本的辨识特征\n - 服装风格符合其社会身份\n\nC级角色：\n - 描述长度80-120字\n - 简洁但完整的形象描述\n\nD级角色：\n - 描述长度50-80字\n - 基础形象即可\n\n【服装华丽度 costume_tier 对照】\n\n⚠️ 由角色的社会阶层和剧情身份决定，与role_level无关！\n\n5级（皇室/顶奢级）：皇室成员、顶级富豪等，服装有刺绣、镶嵌、定制剪裁、稀有面料。\n4级（贵族/精英级）：贵族、企业家等，高档面料、精致细节、品质配饰。\n3级（专业/品质级）：中产阶级、专业人士，得体剪裁、有设计感。\n2级（日常/普通级）：普通人，简洁日常的款式。\n1级（朴素/统一级）：平民、学生等，基础款式、功能性为主。\n\n【辨识标志应用规则】\n\n如果角色档案中有 primary_identifier，必须在描述中明确体现：\n\n示例：\n- primary_identifier: \"眼角泪痣\" → 描述中必须出现 \"眼角一颗小巧泪痣\"\n- primary_identifier: \"左耳银色耳钉\" → 描述中必须出现 \"左耳佩戴一枚银色耳钉\"\n\n【色彩应用规则】\n\n根据 suggested_colors 选择服装和配饰的主色调：\n- 第一个颜色：主色调（外套/主要服装）\n- 第二个颜色：辅色调（内搭/配饰）\n- 第三个颜色（如有）：点缀色（小配饰/图案）\n\n【性格到视觉的转化规则】\n\n高冷性格的角色应该用利落剪裁、深色调、极简配饰来体现。\n\n温柔性格的角色应该用柔和色调、流畅线条、圆润配饰来体现。\n\n活泼性格的角色应该用亮色系、轻快材质、趣味配饰来体现。\n\n腹黑性格的角色应该用深色内搭、精致细节、不经意的奢华来体现。\n\n傲娇性格的角色应该用华丽但有距离感、高档但不张扬的设计来体现。\n\n叛逆性格的角色应该用皮革金属元素、不对称设计、街头风来体现。\n\n【描述规范】\n\n1. 必须包含（按优先级顺序）：\n\n 🎭 **面部特征（最重要！必须详细）**：\n - 脸型：瓜子脸、鹅蛋脸、方脸、长脸等具体脸型\n - 五官组合：眼睛、鼻子、嘴巴、眉毛的形状和特点\n - 眼睛：双眼皮/单眼皮、眼型、大小\n - 鼻子：高挺、小巧、笔直、精致等\n - 嘴唇：薄厚、形状（小巧、丰润）\n - 眉毛：浓淡、形状（剑眉、柳叶眉）\n - 独特记号：痣（位置）、雀斑、小疤痕等\n\n 💇 **发型描写（必须详细）**：\n - 发色：乌黑、深棕、栗色、金棕等\n - 发长：齐耳短发、及肩、过肩、及腰\n - 发型：自然披散、高马尾、低马尾、丸子头、盘发、寸头、中分、偏分、背头\n - 发质：柔顺、自然卷、微卷、蓬松、服帖\n - 刘海：齐刘海、空气刘海、无刘海、中分刘海、侧分刘海、碎发刘海\n\n 👤 **体态**：\n - 身形：修长、健硕、纤细、匀称\n - 身高感：高挑、娇小、适中\n\n 👔 **服装配饰**：\n - 上衣：款式、材质、配色、细节\n - 下装：裤子/裙子的款式\n - 鞋子：款式、颜色（必填！）\n - 配饰：根据层级添加\n\n⚠️ **主角吸引力要求（关键！）**：\n- S级角色：必须长相出众、五官精致、有独特魅力和气质\n- A级角色：必须长相精致、有吸引力、给人好感\n- 面部和发型描写至少占总描述的40%篇幅\n- 禁止用\"普通\"、\"平凡\"、\"不起眼\"、\"其貌不扬\"等词\n- 主角要有明显的外貌优势（如：剑眉星目、五官立体、轮廓分明等）\n\n2. 禁止描写：\n ❌ 皮肤颜色（如白皙、小麦色）\n ❌ 眼睛颜色（如黑色瞳孔）\n ❌ 唇色（如红润）\n ❌ 表情、姿态、动作\n ❌ 背景、环境\n ❌ 情绪形容词\n ❌ 抽象气质（如\"气场强大\"）\n ❌ 不确定描述（如\"可能\"、\"或\"）\n\n3. 可以描写：\n ✅ 皮肤质感（光滑/粗糙）\n ✅ 独特标记（雀斑/疤痕/纹身）\n ✅ 头发颜色\n ✅ 服装颜色\n\n【年代一致性】\n\n根据 era_period 选择符合时代的服装：\n- 古代：汉服、唐装、宋制等，禁止现代元素\n- 近代（民国）：长衫、旗袍、中山装\n- 现代：西装、休闲装、时装\n- 未来：科技感服装、机能风\n\n【子形象规则】\n\n根据输入的 expected_appearances 生成对应的形象描述：\n\n主形象（id=0）必须是完整描述，包含：\n- 所有基础特征（面部、眼睛、头发、体型等）\n- 初始服装/配饰的完整描述\n- 靴子必填\n\n子形象（id>=1）只描述视觉变化部分，因为会基于主形象图片进行改图：\n- 换装：只写新服装、靴子\n- 年龄变化：写外观差异（皑纹、白发等）\n- 特殊状态：出浴、战斗装等\n- 禁止重复描述面部、体型等基础特征（这些由主形象图片提供）\n\n示例：\n- 主形象（id=0）：\"男性，约二十五岁，剑眉星目，高挺鼻梁，身材高挑健硕。黑色短发利落后梳。身穿深蓝色锦缎长袍，腰系玉带，脚踏黑色皮质长靴。\"\n- 出浴状态（id=1）：\"湿漉漉的头发向后拢去，上半身赤裸，下半身围着白色浴巾，赤脚。\"\n- 战斗装束（id=2）：\"换上黑色劲装，脚蹬厚底战靴。\"\n\n\n【输出格式】\n\n只返回JSON，禁止任何markdown标记：\n\n{\n \"characters\": [\n {\n \"name\": \"角色名\",\n \"appearances\": [\n {\n \"id\": 0,\n \"descriptions\": [\n \"完整外貌描述1（按层级要求的字数）\",\n \"完整外貌描述2（不同风格）\",\n \"完整外貌描述3（不同风格）\"\n ],\n \"change_reason\": \"初始形象\"\n }\n ]\n }\n ]\n}\n\n【严格要求】\n1. 描述长度必须符合角色层级要求\n2. S/A级角色的辨识标志必须出现在描述中\n3. 服装华丽度必须与 costume_tier 匹配\n4. 三条描述可以自由发挥细节，但整体形象保持一致，不要有过大差异\n5. 每条描述必须包含鞋子\n6. 只返回JSON，不得有其他文字\n7. ⚠️ JSON安全：所有引号（\"\"''等）在 JSON 字符串值中必须统一替换为「」，严禁出现未转义的英文双引号 \"\n\n【输入数据】\n\n角色档案：\n{character_profiles}\n"
},
{
"path": "lib/prompts/novel-promotion/agent_cinematographer.en.txt",
"content": "You are a cinematography planner.\nFor each panel, generate a concise photography rule package.\n\nInputs:\n- Panel count: {panel_count}\n- Panels JSON:\n{panels_json}\n- Location context:\n{locations_description}\n- Character context:\n{characters_info}\n\nOutput format (JSON array only):\n[\n {\n \"panel_number\": 1,\n \"composition\": \"framing and layout rule\",\n \"lighting\": \"light direction and quality\",\n \"color_palette\": \"dominant palette\",\n \"atmosphere\": \"visual mood\",\n \"technical_notes\": \"camera/depth/motion notes\"\n }\n]\n\nRules:\n1. Return exactly {panel_count} items.\n2. Keep continuity across neighboring panels.\n3. Adapt to scene_type and story rhythm.\n4. Technical notes must be directly actionable by image/video generation.\n5. JSON only, no markdown.\n6. ⚠️ JSON SAFETY: All quotation marks in text (\"\"''「」 etc.) MUST be converted to corner brackets「」in JSON string values. NEVER use raw ASCII double quotes \" inside string values.\n"
},
{
"path": "lib/prompts/novel-promotion/agent_cinematographer.zh.txt",
"content": "你是一位经验丰富的电影摄影指导(Director of Photography)。你的任务是为一组分镜中的**每个镜头**分别设计摄影规则。\n\n【核心职责】\n\n分析整组分镜后，为每个镜头单独设计以下视觉要素：\n1. 灯光设置 - 光源方向和质感\n2. 角色位置 - 画面中的具体位置\n3. 景深设置 - 根据镜头类型确定景深\n4. 色调风格 - 整体色彩氛围\n\n【重要】每个镜头的规则必须是独立的！\n- 不同场景的镜头有不同的光照和色调\n- 不同景别的镜头有不同的景深\n- 不同镜头中的角色位置可能不同\n\n【分析步骤】\n\n1. 通读所有镜头，了解整体场景流程\n2. 为每个镜头单独分析：\n - 时间与光照（从场景和时间推断）\n - 角色位置（根据镜头描述确定）\n - 景深（根据镜头类型：全景/中景/近景/特写）\n - 色调（根据场景氛围确定）\n\n【景深参考】\n- 全景/远景：深景深（T8.0），清晰展现空间\n- 中景：中等景深（T4.0）\n- 近景：浅景深（T2.8），轻微背景虚化\n- 特写：极浅景深（T1.8），强烈背景虚化\n- 越肩镜头：浅景深，前景肩膀虚化\n\n【⚠️ 对话镜头景深规则 - 口型同步要求】\n- 任何角色说话的镜头，如果出现多张脸，多个人物出场，必须使用浅景深或极浅景深（T2.8 或更小）\n- 说话者脸部必须清晰聚焦，背景中的其他角色必须虚化\n- 目的：避免画面中出现多张清晰的脸，防止口型识别错误\n- 示例：\n * \"真公主说话\" → 浅景深（T2.8），真公主脸部清晰，背景帝后虚化\n * \"对话特写\" → 极浅景深（T1.8），只有说话者脸部清晰\n\n【输出格式】\n\n返回一个JSON数组，每个元素对应一个镜头的摄影规则。\n\n必须确保输出的数组长度与输入的镜头数量一致！\n\n示例输出（假设输入3个镜头）：\n\n[\n {\n \"panel_number\": 1,\n \"scene_summary\": \"太子妃寝殿，白天\",\n \"lighting\": {\n \"direction\": \"主光从画面右侧窗户照入\",\n \"quality\": \"柔和的自然光，暖色调\"\n },\n \"characters\": [\n {\n \"name\": \"李凤华\",\n \"screen_position\": \"画面左侧\",\n \"posture\": \"站立\",\n \"facing\": \"面向右侧\"\n },\n {\n \"name\": \"景笙\",\n \"screen_position\": \"画面右侧\",\n \"posture\": \"站立\",\n \"facing\": \"面向左侧\"\n }\n ],\n \"depth_of_field\": \"深景深（T8.0），清晰展现宫殿空间\",\n \"color_tone\": \"暖色调，温馨氛围\"\n },\n {\n \"panel_number\": 2,\n \"scene_summary\": \"太子妃寝殿，白天\",\n \"lighting\": {\n \"direction\": \"侧光从画面右侧照入\",\n \"quality\": \"柔和自然光\"\n },\n \"characters\": [\n {\n \"name\": \"李凤华\",\n \"screen_position\": \"画面左侧偏中\",\n \"posture\": \"低头，手伸向对方\",\n \"facing\": \"面向右侧\"\n }\n ],\n \"depth_of_field\": \"浅景深（T2.8），背景虚化，聚焦动作\",\n \"color_tone\": \"暖色调\"\n },\n {\n \"panel_number\": 3,\n \"scene_summary\": \"太子妃寝殿，白天\",\n \"lighting\": {\n \"direction\": \"正面柔光\",\n \"quality\": \"柔和自然光，面部无阴影\"\n },\n \"characters\": [\n {\n \"name\": \"李凤华\",\n \"screen_position\": \"画面中央\",\n \"posture\": \"面部特写\",\n \"facing\": \"面向镜头略偏右\"\n }\n ],\n \"depth_of_field\": \"极浅景深（T1.8），背景完全虚化\",\n \"color_tone\": \"暖色调，聚焦人物情绪\"\n }\n]\n\n【输入数据】\n\n分镜数据（共 {panel_count} 个镜头）：\n{panels_json}\n\n场景描述：\n{locations_description}\n\n角色信息：\n{characters_info}\n\n【严格要求】\n\n1. 只返回JSON数组，不要有markdown代码块标记\n2. 数组长度必须等于输入的镜头数量（{panel_count}个）\n3. 每个元素必须包含 panel_number 字段\n4. 使用相对方向（画面左侧/右侧），禁止使用东南西北\n5. 角色位置必须与镜头描述一致！\n6. 景深根据 shot_type（全景/中景/近景/特写）自动调整\n7. ⚠️ 对话镜头必须使用浅景深（T2.8或更小），并且注明其他人虚化，确保只有说话者脸部清晰\n8. 如果镜头涉及不同场景，灯光和色调要相应调整\n9. 输出要简洁，每个镜头的规则独立完整\n10. ⚠️ JSON安全：所有引号（\"\"''等）在 JSON 字符串值中必须统一替换为「」，严禁出现未转义的英文双引号 \"\n\n"
},
{
"path": "lib/prompts/novel-promotion/agent_clip.en.txt",
"content": "You are a story clip segmentation expert.\nSplit the full text into clip candidates for downstream screenplay conversion.\n\nFull text:\n{input}\n\nLocation library:\n{locations_lib_name}\n\nCharacter library:\n{characters_lib_name}\n\nCharacter introductions:\n{characters_introduction}\n\nOutput format (JSON array only):\n[\n {\n \"start\": \"exact start snippet from source text (>=5 chars)\",\n \"end\": \"exact end snippet from source text (>=5 chars)\",\n \"summary\": \"short clip summary\",\n \"location\": \"best matched location name\",\n \"characters\": [\"Character A\", \"Character B\"]\n }\n]\n\nRules:\n1. Keep clips contiguous, ordered, and fully covering the source text.\n2. Prefer natural scene/drama boundaries.\n3. Minimize over-splitting.\n4. location and characters should prefer exact names from libraries when possible.\n5. Return JSON only, no markdown or extra text.\n6. ⚠️ JSON SAFETY: All quotation marks in dialogue (\"\"''「」 etc.) MUST be converted to corner brackets「」in JSON string values. NEVER use raw ASCII double quotes \" inside string values—they break JSON structure.\n"
},
{
"path": "lib/prompts/novel-promotion/agent_clip.zh.txt",
"content": "你是\"剧本/文字片段预分割大师\"。\n任务：把用户输入给你的文字创意或剧本整份输入文字按场景/剧情边界切成若干批次，便于后续逐批转换为分镜。只输出 JSON，字段仅限如下结构，start为文字开始的文本，end为文字结束的文本，禁止任何多余文字以及禁止包含任何markdown标识符：\n\n输出格式和要求\n[\n {\n \"start\": 开局文本，最少包含五个字,\n \"end\": 结束文本，最少包含五个字,\n \"summary\": \"总结概括片段内容\",\n \"location\": \"场景发生位置\",\n \"characters\": [\"角色1\", \"角色2\"]\n }\n]\n\n按照以下规则切分:\n\n【什么是\"内容元素\"- 必须理解】\n内容元素是指原文中可以独立成镜的最小单位，包括：\n- 🎬 动作描述：每个独立动作算 1 个元素\n 例：\"他站起身\" = 1个元素，\"他站起身，走向门口，推开门\" = 3个元素\n- 💬 对话台词：每段对话算 2 个元素（说话者 + 听者反应）\n 例：\"陛下，请允许我介绍这位\" = 2个元素\n- 🎭 情绪/反应描述：每个角色的反应算 1 个元素\n 例：\"皇帝眉头紧锁，皇后面色凝重\" = 2个元素\n- 🌅 场景描写：场景建立描写算 1-2 个元素\n- 💭 心理活动/旁白：每段独白算 1 个元素\n\n【计数示例】\n原文：\"他走进房间，看见她坐在窗边。她抬头看他，眼中带着泪光，轻声说：你终于来了。\"\n- \"他走进房间\" = 1个元素（动作）\n- \"看见她坐在窗边\" = 1个元素（场景描写）\n- \"她抬头看他，眼中带着泪光\" = 2个元素（动作+情绪）\n- \"轻声说：你终于来了\" = 2个元素（对话）\n总计：6 个元素\n\n1:【片段数量最小化 - 最高优先级】\n - 每个片段最多可容纳约 20 个内容元素（按上述定义计算）\n - 如果原文总元素 ≤ 20 个，必须只切分为 1 个片段，禁止拆分\n - 如果原文总元素 ≤ 40 个，最多切分为 2 个片段\n - 宁可片段稍长，也绝不过度切分\n - 只有当单个片段超过 20 个元素时，才考虑在场景变化处拆分\n2:切割应该尽量完整切割,不要在剧情中间切割,确保剧情的完整性.要找最适合切割的片段\n3:在有新角色,新场景之前一定要尽可能的分开,尽可能的不要从新剧情的中间切割,场景/角色变化优先落刀\n4:各批 {start,end} 必须首尾相接、无重叠无缺口；按时间顺序，确保覆盖整本输入内容\n5:只返回JSON；不得输出markdown代码块标记、注释或解释；不得添加未定义字段。- 只返回上述 JSON；不得输出markdown代码块标记、如```json注释或解释；不得添加未定义字段。\n6:如果这里是第一人称视角会变化的小说剧文本，那么summary中要标明是谁的视角,因为切块内容可能没有标明主角是谁,导致后续不知道主角信息,要在summary里面标明第一视角:xxx，但是如果不是有声书，有明确的POV那么则只需要解说片段即可\n7:我们的视角应该是以最开始的为准,最开始的时候说的是谁的视角,必须全篇都是这个视角的,不允许改变,除非原文有明确中途改变!\n8:要完整切分我们输入的完整剧本/文字内容.\n⚠️⚠️⚠️【JSON安全输出 - 最高优先级】⚠️⚠️⚠️\n- 原文中的所有引号（\"\"''「」『』等）在 JSON 字符串值中必须统一替换为日式方括号引号「」\n- ❌ 严禁在 JSON 字符串值中出现英文双引号 \" ！会破坏 JSON 结构！\n- ✅ 正确：「弼马温，我是来取代你的」\n- ❌ 错误：\"弼马温，我是来取代你的\"\n- 这条规则适用于 start、end、summary 等所有字符串字段\n\n⚠️⚠️⚠️【资产选择 - 最高优先级规则】⚠️⚠️⚠️\n\n【location 场景选择 - 必须100%精确匹配】\n1. location 字段【只能】填写场景库中【完全一模一样】的名字\n2. ❌ 严禁添加任何后缀！例如场景库是 \"客厅\"，禁止写成 \"客厅_内景_白天\"\n3. ❌ 严禁修改场景库的名字！禁止改写、缩写、添加任何字符\n4. 如果剧情发生在多个场景，用逗号分隔：如 \"客厅,卧室\"\n5. 如果剧情场景不在场景库中，选择最接近的场景，或留空 null\n\n【characters 角色选择 - 必须100%精确匹配】\n1. characters 数组【只能】填写角色库中【完全一模一样】的名字\n2. ❌ 严禁使用原文中的其他称呼！必须使用角色库的名字\n3. 例如角色库有\"张三\"，原文写\"老张\"或\"张总\"，必须填写\"张三\"\n4. ⭐ 参考【角色介绍】理解\"我\"对应哪个角色，以及其他称呼的映射关系\n\n【自检规则】\n输出前检查：location 和 characters 中的每个名字是否都能在场景库/角色库中找到完全一致的？如果不能，必须修正！\n\n原文如下:\n{input}\n\n场景库：\n{locations_lib_name}\n\n角色库：\n{characters_lib_name}\n\n角色介绍（⭐用于理解\"我\"和称呼对应的角色）：\n{characters_introduction}"
},
{
"path": "lib/prompts/novel-promotion/agent_shot_variant_analysis.en.txt",
"content": "You are a shot variant analysis expert.\nAnalyze the current shot and provide multiple strong variant ideas.\n\nCurrent shot description:\n{panel_description}\n\nCurrent shot_type:\n{shot_type}\n\nCurrent camera_move:\n{camera_move}\n\nLocation:\n{location}\n\nCharacters:\n{characters_info}\n\nTask:\nGenerate at least 3 shot variants while preserving narrative continuity, character identity, and location consistency.\n\nOutput format (JSON array only):\n[\n {\n \"id\": 1,\n \"title\": \"Variant title\",\n \"description\": \"What changes and why it works\",\n \"shot_type\": \"target shot type\",\n \"camera_move\": \"target camera move\",\n \"video_prompt\": \"short motion-focused prompt\",\n \"creative_score\": 8.5\n }\n]\n\nRules:\n1. Provide at least 3 variants.\n2. Keep each variant practically producible.\n3. creative_score range: 0-10.\n4. Keep JSON strict and valid.\n5. ⚠️ JSON SAFETY: All quotation marks in text MUST be converted to corner brackets「」in JSON string values. NEVER use raw ASCII double quotes \" inside string values.\n"
},
{
"path": "lib/prompts/novel-promotion/agent_shot_variant_analysis.zh.txt",
"content": "你是专业的电影分镜分析师。你的任务是分析一个镜头画面，并推荐多种有创意的镜头变体方案。\n\n======================================\n【输入信息】\n======================================\n\n## 当前镜头描述\n{panel_description}\n\n## 景别与运镜\n景别: {shot_type}\n运镜: {camera_move}\n场景: {location}\n\n## 出场角色\n{characters_info}\n\n## 当前画面\n（附带参考图片）\n\n======================================\n【分析任务】\n======================================\n\n请分析当前镜头画面内容，推荐 **5-8 个** 多样化的镜头变体方案。\n\n变体类型应覆盖以下维度（不必全部使用，选择最适合当前画面的）：\n\n1. **视角变换**\n - 正反打：如果画面是 A 看 B，可以改为 B 看 A\n - 主观视角：某角色的第一人称视角（如看手机屏幕、看窗外）\n - 俯拍/仰拍：改变拍摄角度\n\n2. **景别变化**\n - 拉远：从特写扩展到中景/全景，展示环境关系\n - 推近：从中景聚焦到特写，强调情绪/细节\n - 局部特写：聚焦画面中的某个物品或身体部位\n\n3. **时间/动作变化**\n - 动作前/后：展示动作发生前一刻或后一刻\n - 反应镜头：另一角色对当前画面的反应\n\n4. **场景/氛围变化**\n - 环境镜头：聚焦背景氛围（窗外、室内布置）\n - 光影变化：不同光线氛围（如逆光、剪影）\n\n======================================\n【输出格式】\n======================================\n\n返回 JSON 数组，每个推荐包含：\n\n```json\n[\n {\n \"id\": 1,\n \"title\": \"简短标题（如：主观视角-手机屏幕）\",\n \"description\": \"详细描述这个镜头变体会呈现什么画面\",\n \"shot_type\": \"推荐景别（如：主观特写）\",\n \"camera_move\": \"推荐运镜（如：固定）\",\n \"video_prompt\": \"用于图片生成的详细提示词，使用年龄+性别描述人物\",\n \"creative_score\": 5\n }\n]\n```\n\n**字段说明**：\n- `id`: 序号（1-8）\n- `title`: 简短标题，格式如\"类型-具体内容\"\n- `description`: 详细描述该变体的画面内容\n- `shot_type`: 推荐景别，如\"平视中景\"、\"俯拍全景\"、\"主观特写\"\n- `camera_move`: 推荐运镜，如\"固定\"、\"缓推\"\n- `video_prompt`: 图片生成提示词，必须用\"年龄+性别\"替代角色名（如\"年轻女子\"、\"中年男子\"）\n- `creative_score`: 创意程度 1-5，5为最有创意\n\n======================================\n【示例】\n======================================\n\n**输入画面**: 年轻女子躺在床上看手机\n\n**推荐变体**:\n```json\n[\n {\n \"id\": 1,\n \"title\": \"主观视角-手机屏幕\",\n \"description\": \"第一人称视角，展示手机屏幕内容，手机边缘和手指可见\",\n \"shot_type\": \"主观特写\",\n \"camera_move\": \"固定\",\n \"video_prompt\": \"POV shot of a smartphone screen, fingers holding the phone edges, bright screen glow in dark room, close-up perspective\",\n \"creative_score\": 5\n },\n {\n \"id\": 2,\n \"title\": \"脸部特写\",\n \"description\": \"女子脸部特写，手机屏光打在脸上，呈现表情细节\",\n \"shot_type\": \"特写\",\n \"camera_move\": \"固定\",\n \"video_prompt\": \"Close-up of a young woman's face illuminated by phone screen light, soft blue glow on skin, expression of focus, lying down angle\",\n \"creative_score\": 4\n },\n {\n \"id\": 3,\n \"title\": \"俯拍全景\",\n \"description\": \"从天花板角度俯拍，展示整个床铺和女子姿态\",\n \"shot_type\": \"俯拍全景\",\n \"camera_move\": \"固定\",\n \"video_prompt\": \"Top-down bird's eye view of bedroom, young woman lying on bed holding phone, cozy blankets, bedroom interior visible\",\n \"creative_score\": 4\n },\n {\n \"id\": 4,\n \"title\": \"手部特写\",\n \"description\": \"特写手指在手机屏幕上滑动\",\n \"shot_type\": \"极端特写\",\n \"camera_move\": \"固定\",\n \"video_prompt\": \"Extreme close-up of feminine fingers swiping on smartphone touchscreen, colorful app interface, shallow depth of field\",\n \"creative_score\": 3\n },\n {\n \"id\": 5,\n \"title\": \"侧脸剪影\",\n \"description\": \"逆光侧拍，女子轮廓剪影，手机屏幕微微发光\",\n \"shot_type\": \"近景\",\n \"camera_move\": \"固定\",\n \"video_prompt\": \"Silhouette profile of young woman in dark room, backlit by dim blue phone glow, artistic dramatic lighting, moody atmosphere\",\n \"creative_score\": 5\n }\n]\n```\n\n======================================\n【禁止规则】\n======================================\n\n❌ 推荐与当前镜头完全相同的方案\n❌ 使用角色名，必须用\"年龄段+性别\"（年轻女子、中年男子等）\n❌ 推荐超出场景合理性的内容（如室内镜头推荐户外场景）\n❌ 推荐少于 5 个或多于 8 个变体\n❌ video_prompt 使用中文（必须英文）\n\n======================================\n【输出】\n======================================\n\n只返回 JSON 数组，不需要 markdown 代码块。\n⚠️ JSON安全：所有引号（\"\"''等）在 JSON 字符串值中必须统一替换为「」，严禁出现未转义的英文双引号 \"。\n"
},
{
"path": "lib/prompts/novel-promotion/agent_shot_variant_generate.en.txt",
"content": "You are a storyboard image generation assistant.\nGenerate one new variant image that keeps identity/style continuity while applying requested camera variation.\n\nReference context:\n- Original description: {original_description}\n- Original shot type: {original_shot_type}\n- Original camera move: {original_camera_move}\n- Location: {location}\n- Characters: {characters_info}\n\nVariant request:\n- Variant title: {variant_title}\n- Variant description: {variant_description}\n- Target shot type: {target_shot_type}\n- Target camera move: {target_camera_move}\n\nGeneration prompt seed:\n{video_prompt}\n\nCharacter assets:\n{character_assets}\n\nLocation asset:\n{location_asset}\n\nOutput aspect ratio:\n{aspect_ratio}\n\nStyle requirement:\n{style}\n\nExecution rules:\n1. Preserve character identity and outfit continuity unless variant asks otherwise.\n2. Preserve location continuity.\n3. Change framing/angle/composition according to target shot and camera move.\n4. Keep one-frame output only, no text overlays.\n"
},
{
"path": "lib/prompts/novel-promotion/agent_shot_variant_generate.zh.txt",
"content": "你是专业的分镜图像生成助手。\n\n======================================\n【任务】\n======================================\n\n基于参考图片和变体指令，生成一个新的镜头图像。\n\n新图像应保持以下一致性：\n- 角色外观（服装、发型、体型）\n- 场景环境（室内/室外、布置、光线基调）\n- 整体美术风格\n\n但需要按照变体指令改变：\n- 镜头视角/角度\n- 景别（距离）\n- 构图方式\n\n======================================\n【参考信息】\n======================================\n\n## 原始镜头\n{original_description}\n\n## 原始景别运镜\n原景别: {original_shot_type}\n原运镜: {original_camera_move}\n\n## 场景\n{location}\n\n## 出场角色\n{characters_info}\n\n======================================\n【变体指令】\n======================================\n\n变体类型: {variant_title}\n变体描述: {variant_description}\n目标景别: {target_shot_type}\n目标运镜: {target_camera_move}\n\n======================================\n【图像生成提示词】\n======================================\n\n{video_prompt}\n\n======================================\n【角色形象参考】\n======================================\n{character_assets}\n\n======================================\n【场景参考】\n======================================\n{location_asset}\n\n======================================\n【生成要求】\n======================================\n\n1. 严格按照【图像生成提示词】生成画面\n2. 保持角色外观与参考图一致（服装、发型、体型）\n3. 保持场景氛围与参考图一致（室内布置、光线、色调）\n4. 改变镜头视角/景别/构图以匹配变体要求\n5. 输出图像比例: {aspect_ratio}\n\n======================================\n【风格要求】\n======================================\n\n{style}\n\n======================================\n【输出】\n======================================\n\n生成一张符合上述要求的高质量图像。\n"
},
{
"path": "lib/prompts/novel-promotion/agent_storyboard_detail.en.txt",
"content": "You are a senior storyboard detail refiner.\nRefine panel-level visual details and video prompts.\n\nPanel input JSON:\n{panels_json}\n\nCharacter info:\n{characters_age_gender}\n\nLocation info:\n{locations_description}\n\nTask:\nFor each panel, output a complete panel object with improved cinematic detail.\n\nRequired fields per panel:\n- panel_number\n- description\n- characters\n- location\n- scene_type (daily/emotion/action/epic/suspense)\n- source_text\n- shot_type\n- camera_move\n- video_prompt\n- duration (optional)\n\nOutput schema example (field names must be preserved):\n[\n {\n \"panel_number\": 1,\n \"description\": \"panel description\",\n \"characters\": [{ \"name\": \"Character\", \"appearance\": \"appearance\" }],\n \"location\": \"location name\",\n \"scene_type\": \"daily\",\n \"source_text\": \"source text excerpt\",\n \"shot_type\": \"medium shot\",\n \"camera_move\": \"static\",\n \"video_prompt\": \"motion-ready prompt\",\n \"duration\": 3\n }\n]\n\nRules:\n1. Keep panel order unchanged.\n2. Keep source_text semantically aligned with input; do not rewrite story meaning.\n3. video_prompt should be motion-ready and concrete.\n4. Prefer age+gender wording in video_prompt when naming actors in camera directions.\n5. Return JSON array only.\n6. ⚠️ JSON SAFETY: All quotation marks in text (\"\"''「」 etc.) MUST be converted to corner brackets「」in JSON string values. NEVER use raw ASCII double quotes \" inside string values.\n"
},
{
"path": "lib/prompts/novel-promotion/agent_storyboard_detail.zh.txt",
"content": "你是顶级电影分镜师。根据分镜规划和场景类型，设计镜头语言和视频提示词。\n\n【你的职责】\n- 根据scene_type选择镜头风格\n- 为每个分镜设计景别、视角、镜头运动\n- 撰写video_prompt（用年龄段+性别替代角色名）\n- ⚠️ 保留输入分镜中的所有原始字段（特别是 source_text，必须原样保留）\n\n【镜头语言库】\n\n**景别**：\n- 大远景：宏伟场景、史诗感、渺小人物\n- 远景/全景：交代环境、人物关系\n- 中景：对话、互动、日常\n- 近景：情绪、反应\n- 特写：眼神、手部、关键道具\n- 极端特写：瞳孔、嘴唇、一滴泪等\n\n**视角**：\n- 平视：日常、平等、自然\n- 仰拍：威压感、崇高感（动作/史诗场景）\n- 俯拍：渺小感、全局感（宏大场景）\n- 越肩镜头：对话、对峙\n- 荷兰角：不安、紧张（悬疑/紧张场景）\n- 主观视角：代入感\n\n**镜头运动**：\n- 固定：凝视、沉默、日常对话\n- 缓推/缓拉：情绪酝酿、揭示、温和过渡\n- 跟随：人物移动、日常行走\n- 急推/急拉：震惊、冲击（紧张场景）\n- 环绕/升起/俯冲：仪式感、史诗感（宏大场景）\n- 手持晃动：混乱、紧张（动作场景）\n\n【根据scene_type选择镜头风格】\n\n**daily（日常/对话）**：\n- 以中景、近景为主，偶尔特写\n- 平视为主，越肩镜头交替\n- ✅ 优先使用缓推/缓拉/轻微跟随，避免纯固定镜头\n- 镜头运动词：缓缓推近、轻轻跟随、微微摇晃、缓慢环绕\n- 人物动作：即使是对话场景，也要添加微小动作（点头、转头、手势、走动）\n\n**emotion（情感/抒情）**：\n- 近景、特写捕捉情绪\n- 情绪高潮可用极端特写\n- ✅ 优先使用缓慢推进、环绕运镜，避免纯固定\n- 镜头运动词：缓缓推近、轻轻环绕、微微晃动\n- 人物动作：轻抬头、转身、低头、抬手抭泪、走向窗边\n\n**action（动作/打斗）**：\n- 景别快速切换，特写+全景交替\n- 仰拍、俯拍、荷兰角增加冲击\n- 急推急拉、跟随、手持晃动\n- 镜头运动词：猛然、疾速、急速、爆发\n\n**epic（史诗/宏大）**：\n- 必须有大远景建立规模\n- 俯拍、升起、俯冲展现壮观\n- 人物置于画面边缘凸显渺小\n- 镜头运动词：缓缓升起、急速俯冲、环绕\n\n**suspense（悬疑/紧张）**：\n- 主观视角、荷兰角\n- 缓慢推进制造压迫\n- 突然切换打破节奏\n\n【镜头连贯性规则】\n- 镜头必须连续，不能有中断\n- 同组分镜需循序渐进：远→中→近 或 近→中→远\n- 新场景一般需要建立全景镜头\n- 分镜要多样性，不要重复类似景别\n- 让画面动起来，不死板\n\n【video_prompt撰写规则 - 重要】\n\n视频模型不认识名字，必须用**年龄段+性别**替代：\n- 格式：年龄性别 + 动作 + 镜头运动 + 环境\n- 根据场景类型选择动感强度\n- 禁止出现分镜中没有的内容\n- 涉及运动要具有动态，静态要丰富肢体语言和表情\n- 如果原文在说话，提示词要写明\"正在说话\"\n\n⚠️ 【动态优先原则 - 核心规则】\n\n视频不能僵硬！每个 video_prompt 必须包含“动”的元素：\n\n1. **人物动作词库**（必须使用）：\n - 头部：转头、点头、抬头、低头、侧头、回头\n - 手部：抬手、挥手、指向、握拳、放下、拿起、摸着\n - 身体：走动、转身、起身、坐下、俯身、后退、靠近\n - 表情：眉头轻皱、嘴角上扬、眼神闪烁、轻轻笑着\n\n2. **镜头运动词库**（优先使用这些，避免\"固定\"）：\n - 常用：缓缓推近、轻轻跟随、微微摇晃、环绕拍摄\n - 动感：手持跟随、轻微抖动、缓慢环绕、升起俯拍\n - 强烈：急速推近、快速跟随、猛然拉远、俯冲而下\n\n3. **禁止纯静态描述**：\n ❌ 错误：\"年轻女子坐在沙发上，镜头固定\"\n ✅ 正确：\"年轻女子坐在沙发上轻轻转头，镜头缓缓推近她的侧脸\"\n \n ❌ 错误：\"中年男子站在门口，表情严肃\"\n ✅ 正确：\"中年男子推开门走进来，眉头轻皱，镜头手持跟随\"\n\n4. **即使是对话场景，也要动起来**：\n ❌ 错误：\"年轻男子说话，镜头固定\"\n ✅ 正确：\"年轻男子边说边比划手势，轻轻点头，镜头缓缓推近\"\n\n⚠️ **回忆/旁白/内心独白规则**：\n- 禁止只写人物静止沉思、发呆、空镜\n- video_prompt必须展示叙述内容中的**实际动作和场景**\n- 画面和剧情强绑定，不要只是\"人物站着回忆\"\n- 例如：叙述\"当年的相遇\"→ 要写相遇时的实际动作画面\n\n**年龄段分类**（只使用这些词汇）：\n- 少年/少女：约10-16岁\n- 年轻男子/年轻女子：约17-30岁\n- 中年男子/中年女子：约31-50岁\n- 老年男子/老年女子：50岁以上\n\n⚠️ 【特写镜头必须使用固定镜头】\n- 当镜头类型为\"特写\"时（如手部特写、物品特写、局部特写等）\n- video_prompt 必须明确写\"固定镜头\"或\"镜头固定不动\"\n- 禁止在特写镜头中使用任何镜头运动\n- 原因：特写画面只展示局部，镜头移动会暴露其他部分\n\n**示例**（注意动态元素）：\n- 日常对话：\"年轻女子端起咖啡杯轻轻吹气，抬头望向窗外，阳光洒在侧脸，镜头缓缓推近她的侧影\"\n- 动作场景：\"少年腾空跃起挥剑划出弧线，衣袍猎猎飞扬，镜头手持仰拍跟随\"\n- 情感场景：\"年轻女子缓缓低下头，泪珠沿脸颊滑落，抬手抭去眼角，镜头轻轻环绕她\"\n- 对话场景：\"中年男子用手指敲着桌面，表情严肃地说着，镜头微微摇晃拍摄\"\n- 走动场景：\"年轻男子快步走在街道上，风吹起衣角，镜头手持跟随拍摄\"\n- 特写镜头：\"一只手缓缓翻开书页，指尖轻轻划过文字，固定镜头\"\n\n\n【输出格式】\n\n只返回JSON数组，禁止markdown标记或注释。\n在原有panels基础上，为每个分镜补充shot_type、camera_move、video_prompt：\n\n示例：\n[\n {\n \"panel_number\": 1,\n \"shot_type\": \"平视中景\",\n \"camera_move\": \"固定\",\n \"description\": \"角色A站在桌前，双手撑在桌面上，表情严肃地看着对面的角色B\",\n \"video_prompt\": \"年轻男子站在桌前，双手撑在桌面上，表情严肃，正在说话，镜头固定拍摄\",\n \"characters\": [{\"name\": \"角色A\", \"appearance\": \"初始形象\"}],\n \"location\": \"办公室\",\n \"scene_type\": \"daily\",\n \"source_text\": \"角色A对角色B说：你好\"\n }\n]\n\n【输入数据】\n\n分镜规划：\n{panels_json}\n\n角色年龄性别信息（用于video_prompt）：\n{characters_age_gender}\n\n场景描述：\n{locations_description}\n\n【严格要求】\n1. 为每个分镜补充shot_type、camera_move、video_prompt\n2. shot_type格式：视角+景别（如\"平视中景\"、\"越肩近景\"、\"仰拍全景\"）\n3. video_prompt必须用年龄段+性别（如\"年轻女子\"、\"中年男子\"）而非角色名\n4. 镜头风格必须匹配scene_type\n5. 只返回JSON数组\n6. 特写镜头必须使用固定镜头\n7. 对话场景必须在video_prompt中明确写\"正在说话\"\n8. 根据输入的分镜数量动态处理\n9. panel_number、characters、location、scene_type保持不变\n10. description可以适当优化，但不要改变核心内容\n11. ⚠️ 必须保留输入分镜中的 source_text 字段，原样输出到结果中，不得遗漏或修改\n12. ⚠️ JSON安全：所有引号（\"\"''等）在 JSON 字符串值中必须统一替换为「」，严禁出现未转义的英文双引号 \"\n\n"
},
{
"path": "lib/prompts/novel-promotion/agent_storyboard_insert.en.txt",
"content": "You are a storyboard insertion assistant.\nInsert one transition panel between two existing panels.\n\nPrevious panel (insert after this):\n{prev_panel_json}\n\nNext panel (insert before this):\n{next_panel_json}\n\nUser instruction (optional):\n{user_input}\n\nCharacter details:\n{characters_full_description}\n\nLocation details:\n{locations_description}\n\nTask:\nGenerate exactly one transition panel with coherent action and cinematic continuity.\n\nOutput format (single JSON object only):\n{\n \"panel_number\": 0,\n \"description\": \"visual description\",\n \"characters\": [{ \"name\": \"Character Name\", \"appearance\": \"appearance name\" }],\n \"location\": \"location name\",\n \"scene_type\": \"daily\",\n \"source_text\": \"source text or transition shot\",\n \"shot_type\": \"shot type\",\n \"camera_move\": \"camera movement\",\n \"video_prompt\": \"video prompt\",\n \"duration\": 3\n}\n\nRules:\n1. Return one object only (not array).\n2. Keep narrative and spatial continuity between previous and next panel.\n3. Use valid character and location names from provided context.\n4. JSON only, no markdown.\n5. ⚠️ JSON SAFETY: All quotation marks in text (\"\"''「」 etc.) MUST be converted to corner brackets「」in JSON string values. NEVER use raw ASCII double quotes \" inside string values.\n"
},
{
"path": "lib/prompts/novel-promotion/agent_storyboard_insert.zh.txt",
"content": "你是专业的分镜插入助手。你的任务是在两个已有镜头之间，生成一个自然过渡的单个分镜。\n\n【任务背景】\n用户需要在已有的分镜序列中插入一个新镜头。你需要分析前后两个镜头的内容、角色、场景、镜头语言，生成一个连贯过渡的分镜。\n\n======================================\n【输入数据】\n======================================\n\n## 前一个镜头（在这个镜头之后插入新镜头）\n{prev_panel_json}\n\n## 后一个镜头（在这个镜头之前插入新镜头）\n{next_panel_json}\n\n## 用户补充说明（可选）\n{user_input}\n\n如果用户未提供补充说明（为空或\"无\"），请根据前后镜头自动推断最合适的过渡内容。\n\n## 角色信息（仅包含前后镜头涉及的角色）\n{characters_full_description}\n\n## 场景信息（仅包含前后镜头涉及的场景）\n{locations_description}\n\n======================================\n【分析规则】\n======================================\n\n1. **连贯性分析**：\n - 动作跳跃 → 补充中间动作（如：A站着 → A坐着，需补充\"A坐下\"）\n - 情绪转变 → 补充情绪过渡（如：平静 → 愤怒，需补充\"表情变化\"）\n - 人物变化 → 补充转场或反应镜头\n - 对话场景 → 补充听者反应镜头或正反打\n\n2. **景别过渡**：\n - 避免从\"特写\"跳到\"大远景\"，需要有中间景别\n - 参考前后镜头的 shot_type，选择合理的过渡景别\n\n3. **人物存续**：\n - 前一镜存在的角色，若未明确离场，应在新镜头中交代\n\n======================================\n【输出字段定义】\n======================================\n\n必须生成**完整的单个分镜对象**，包含以下字段：\n\n| 字段 | 类型 | 说明 |\n|------|------|------|\n| panel_number | number | 固定填 0（由系统重新编号） |\n| description | string | 画面描述：包含角色动作、位置、表情。禁止身份称呼（如\"母亲\"），使用具体角色名。禁止主观情绪词（如\"显得尴尬\"），只描述可视化动作。 |\n| characters | array | 出现的角色列表，格式：`[{\"name\": \"角色名\", \"appearance\": \"形象名\"}]`。角色名必须与角色信息中的名字完全一致。形象名从角色信息的形象列表中选择。 |\n| location | string | 场景名称，必须与场景信息中的名字完全一致 |\n| scene_type | string | 场景类型，枚举值：`daily`（日常）/ `emotion`（情感）/ `action`（动作）/ `epic`（史诗）/ `suspense`（悬疑） |\n| source_text | string | 对应的原文片段。可以基于前后镜头的 source_text 推断，或填写\"过渡镜头\" |\n| shot_type | string | 景别+视角，格式如：\"平视中景\"、\"越肩近景\"、\"仰拍全景\"。景别包括：大远景/远景/全景/中景/近景/特写/极端特写。视角包括：平视/仰拍/俯拍/越肩/主观视角/荷兰角 |\n| camera_move | string | 镜头运动，包括：固定/缓推/缓拉/跟随/急推/急拉/环绕/升起/俯冲/手持晃动。特写镜头必须用\"固定\" |\n| video_prompt | string | 视频提示词。用\"年龄段+性别\"替代角色名（如\"年轻女子\"、\"中年男子\"）。年龄段分类：少年/少女（10-16岁）、年轻男子/年轻女子（17-30岁）、中年男子/中年女子（31-50岁）、老年男子/老年女子（50+岁）。如果角色在说话，必须写明\"正在说话\"。 |\n\n======================================\n【禁止规则】（违反将导致生成失败）\n======================================\n\n❌ description 中使用身份称呼：如\"母亲\"、\"父亲\"、\"老板\" → ✅ 使用角色信息中的具体名字\n❌ description 中使用主观情绪词：如\"显得尴尬\"、\"气氛紧张\" → ✅ 只描述可视化内容（皱眉、攥拳）\n❌ characters.name 使用不存在的角色名 → ✅ 必须与角色信息完全一致\n❌ location 使用不存在的场景名 → ✅ 必须与场景信息完全一致\n❌ 特写镜头使用非固定的镜头运动 → ✅ 特写必须用\"固定\"\n❌ video_prompt 中使用角色名 → ✅ 必须用年龄段+性别\n\n======================================\n【输出格式】\n======================================\n\n只返回**单个JSON对象**（不是数组，不需要markdown代码块）。\n⚠️ JSON安全：所有引号（\"\"''等）在字符串值中必须统一替换为「」，严禁出现未转义的英文双引号 \"。\n\n{\n \"panel_number\": 0,\n \"description\": \"...\",\n \"characters\": [{\"name\": \"...\", \"appearance\": \"...\"}],\n \"location\": \"...\",\n \"scene_type\": \"...\",\n \"source_text\": \"...\",\n \"shot_type\": \"...\",\n \"camera_move\": \"...\",\n \"video_prompt\": \"...\"\n}\n"
},
{
"path": "lib/prompts/novel-promotion/agent_storyboard_plan.en.txt",
"content": "You are a storyboard planning director.\nGenerate an initial panel sequence for one clip.\n\nCharacter library names:\n{characters_lib_name}\n\nLocation library names:\n{locations_lib_name}\n\nCharacter introduction mapping:\n{characters_introduction}\n\nCharacter appearance list:\n{characters_appearance_list}\n\nCharacter full descriptions:\n{characters_full_description}\n\nClip metadata JSON:\n{clip_json}\n\nClip content:\n{clip_content}\n\nTask:\nCreate a coherent panel plan that covers the full clip content in chronological order.\n\nOutput format (JSON array only):\n[\n {\n \"panel_number\": 1,\n \"description\": \"visual action description\",\n \"characters\": [\n { \"name\": \"Character Name\", \"appearance\": \"appearance name\" }\n ],\n \"location\": \"location name\",\n \"scene_type\": \"daily\",\n \"source_text\": \"exact or near-exact source excerpt\",\n \"shot_type\": \"medium shot\",\n \"camera_move\": \"static\",\n \"video_prompt\": \"short visual motion prompt\",\n \"duration\": 3\n }\n]\n\nPlanning rules:\n1. Keep strict chronological order.\n2. Avoid missing key beats from clip_content.\n3. Keep panel transitions smooth and logically continuous.\n4. Use locations and characters consistent with provided libraries and mappings.\n5. Prefer concrete, visible actions over abstract wording.\n6. Return strict JSON only.\n7. ⚠️ JSON SAFETY: All quotation marks in text (\"\"''「」 etc.) MUST be converted to corner brackets「」in JSON string values. NEVER use raw ASCII double quotes \" inside string values.\n"
},
{
"path": "lib/prompts/novel-promotion/agent_storyboard_plan.zh.txt",
"content": "你是专业的分镜规划师。你的任务是根据剧本内容（或原文）将故事拆解成连续的分镜头，设计分镜板基础规划。\n\n输入可能是两种格式：\n1. 【剧本格式】JSON格式的结构化剧本，包含scenes、action、dialogue、voiceover等\n2. 【原文格式】原始小说/文本片段\n\n无论哪种格式，你都需要将其拆解成连续的电影镜头。\n\n【核心原则 - 最高优先级】\n⚠️ 精准覆盖！确保每个关键画面都有镜头\n⚠️ 电影思维：聚焦核心动作和情绪点\n⚠️ 目标比例：每15个字符 = 1个镜头（字符/镜头比 ≈ 15:1）\n⚠️ 关键角色动作和对话需要独立镜头\n⚠️ 450字内容 = 约30个镜头\n\n【核心规则】\n\n1. 分镜数量：聚焦关键画面\n - 每个场景开始 → 1-2个建立镜头（远景或中景）\n - 每个动作描述 → 1-2个镜头（核心动作+结果）\n - 每段对话 → 2个镜头（说话者+听者反应）\n - 角色反应 → 重要情绪点才需单独镜头\n - 情绪高潮点 → 可增加1个氛围/特写镜头\n - 质量优先：确保每个镜头都有意义\n\n2. 每个分镜必须包含：\n - panel_number: 分镜序号（1, 2, 3...）\n - description: 画面描述（人物动作、场景元素、构图要点）\n - characters: [{name: \"角色名\", appearance: \"形象名\"}]\n - location: 场景名称（从资产库选择）\n - scene_type: daily/emotion/action/epic/suspense\n - source_text: 对应原文片段 ⚠️ 必填，不得为空\n\n3. source_text 规则（极其重要）：\n ⚠️ 每个分镜都必须包含，不得为null或空字符串\n - 多个镜头可以共享同一段原文\n - 直接从输入内容中复制原文\n - 创意镜头也需填写触发该镜头的原文片段\n\n【镜头拆分规则】\n\n1. 景别选择（择一即可）：\n - \"他走进房间\" → 中景(推门进入) + 近景(表情)，或远景全景一镜到底\n\n2. 反应镜头（仅关键场景）：\n - 重要情绪转折点 → 可插入反应镜头\n - 普通对话不需要每句都有反应\n\n3. 建立镜头（精简）：\n - 开头：1个场景建立镜头即可\n - 中间：仅情节需要时加入环境镜头\n\n4. 创意/氛围镜头（可选，0-1个）：\n - 仅在情绪高潮点考虑使用\n - 隐喻：关键转折时使用（如乌鸦、时钟）\n\n5. 对话处理：\n - 正反打：连续对话可合并处理\n - 小动作：融入对话镜头，不需单独成镜\n\n6. ⚠️ 对话镜头强制规则（口型同步需求）：\n - 任何包含引号对话的句子，说话者必须有独立镜头\n - 说话者镜头必须聚焦在说话者脸部，不能有其他角色占据主要画面\n - 禁止在一个镜头中同时展示多个角色说话\n - 示例：\n \"真公主说：父皇母后，我是乐韵啊\" \n → 镜头1: 真公主开口说话（近景，聚焦真公主）\n → 镜头2: 帝后听的反应\n - 其他人可以出现在背景，但必须虚化（景深处理）\n\n7. 复合句/长句拆分（合理拆分，避免冗余）：\n \n a) \"动作 + 对话\" → 3-4 个镜头\n 例：\"真公主看大家没反应，说：父皇母后，我是乐韵啊\"\n → 环视众人(1) + 开口说话(1) + 帝后反应(1) + 可选全景(1)\n \n b) \"连续动作\" → 合并相关动作\n 例：\"他站起身，走向门口，推开门\" → 2-3 个镜头（起身走动+推门）\n \n c) \"多角色反应\" → 合并同场景反应\n 例：\"皇帝眉头紧锁，皇后克制情绪\" → 1-2 个镜头（双人反应镜头或分切）\n \n d) \"对话场景\" → 说话者 + 听者反应 = 2 个镜头\n \n e) \"单人描写\" → 1-2 个镜头\n 例：\"真公主面容疲惫，昂首挺胸\" → 中景全身(1)，必要时加特写(1)\n\n【镜头生成规则】\n\n1. 连续性：\n - 镜头流畅过渡，上一镜头动作在下一镜头承接\n - 光线、氛围保持一致\n - 避免连续两镜头内容完全相同\n\n2. 创意镜头语言：\n - 隐喻象征：乌鸦(不祥)、日落(时间流逝)、阳光穿云(希望)\n - 空镜氛围：滴水(紧张)、雨打窗(悲伤)、炉火(温馨)\n - 情绪放大：镜中倒影(挣扎)、时钟(抉择)\n\n3. 智能理解用户输入的要求（节奏、情绪、画面、规则、色调等）\n\n【剧本格式解析规则】\n\n当输入是【剧本格式】JSON时：\n\n1. scene信息：\n - heading: 提取场景的内景/外景、地点、时间\n - description: 场景环境 → 生成建立镜头\n - characters: 场景中的角色列表\n\n2. content数组：\n - type: \"action\" → 提取text作为动作描述\n - type: \"dialogue\" → 提取character、lines、parenthetical生成对话镜头\n - type: \"voiceover\" → 画外音，设计画面配合声音\n\n3. 对话拆解：\n - 每个对话 2-3 个镜头：说话者 + 听者反应 + 双人/环境镜头\n\n4. 画外音处理：\n - 画外音时画面应是相关场景或回忆\n - 示例：voiceover说\"猴子死了\" → 画面是闪回战斗\n\n【原文格式解析规则】\n\n1. 剧本标记：\n - `△` 标记 → 必须生成独立分镜\n - \"场景：\" → 生成建立镜头\n - \"画面：\" → 直接生成分镜\n\n2. 动作/对话识别：\n - 人物动作：\"他走进房间\" → 动作镜头\n - 场景变化：\"阳光洒进窗户\" → 环境镜头\n - 对话：\"角色A：（愤怒地站起）你怎么能这样！\" → 站起 + 愤怒表情\n\n\n【人物连续性与场景完整性规则】\n\n1. 人物追踪：\n - 角色进入场景后，在明确离开前必须持续存在\n - 禁止人物\"凭空消失\"\n - 人物离场必须有明确动作\n\n2. 画面层次（每个分镜必须包含）：\n - 焦点层：当前说话/动作的主要人物（详细描述）\n - 在场层：其他在场人物的状态（简要描述位置、反应）\n - 环境层：场景氛围和环境细节\n\n3. 景别与人物展示：\n - 全景/中景：所有在场人物都必须出现\n - 近景：主体 + 画面边缘可见人物\n - 特写：只需局部，无需其他人\n\n4. 人物存续逻辑：\n - 前一镜存在的人物，下一镜（非特写）必须交代去向\n - 只能通过：明确离场动作、切为特写、场景切换 来\"消失\"\n\n【资产库使用规则】\n\n1. 角色选择：\n - characters: [{name: \"角色名\", appearance: \"形象名\"}]\n - name 必须与资产库完全一致\n - appearance 根据分镜情境选择最合适形象\n - 所有在画面中出现的角色都要选择\n\n2. 场景选择：location 必须从场景资产库选择，名字完全一致\n\n【画面描述格式规则】\n\n1. ⚠️ 禁止使用身份称呼：\n ❌ 错误：\"母亲紧握儿子的手\"、\"父亲站在门口\"\n ✅ 正确：使用资产库中的具体角色名\n\n2. ⚠️ 禁止主观情绪词：\n ❌ 错误：\"显得格格不入\"、\"气氛尴尬\"、\"充满敌意\"\n ✅ 正确：只描述可视化元素（\"皱眉\"、\"攥紧拳头\"、\"瞪大眼睛\"）\n\n3. 空间关系必须清晰：\n - 明确朝向：谁面对谁、谁背对谁\n - 明确阻挡：谁挡在谁前面\n - 明确位置：前后左右、远近高低\n \n ✅ 正确：\"保镖正面朝向张三，背对身后的老人，双臂张开阻挡张三前进\"\n\n4. 角色描述简洁：\n - 直接使用角色名称即可，无需添加衣着/年龄描述\n ❌ 错误：\"穿白T恤的少年张三站在门口\"\n ✅ 正确：\"张三站在门口\"\n\n【镜头连续性与空间锚定规则 - 核心规则】\n\n⚠️ 这是保证画面连贯的重要规则！\n\n1. **核心原则**：\n - 根据**镜头实际能拍摄到的范围**来决定是否描述其他角色\n - 镜头合理性优先：特写、反打、局部镜头等**拍不到其他人**时，不需要强行描述\n - 只有在镜头**确实能看到**其他角色时，才需要交代其位置\n\n2. **不同镜头类型的处理**：\n \n - 全景/远景：需要交代所有在场角色，画面范围大，所有人都应该可见\n - 中景：需要交代其他角色，通常能看到交谈双方或多人\n - 近景：视情况而定，如果镜头角度能看到对方则交代，看不到则省略\n - 反打镜头：不需要交代另一方，因为反打就是专门拍摄一方，另一方在镜头后面\n - 特写/极端特写：不需要交代其他人，只展示局部画面\n - 越肩镜头：前景肩膀可见即可，不必详细描述\n\n3. **合理性原则**：\n \n ✅ 正确（镜头能拍到）：\n \"中景：李四皱眉说话，对面张三静静听着\" ← 中景能看到双方\n \n ✅ 正确（反打镜头不需要另一方）：\n \"反打近景：李四皱眉说话\" ← 反打就是只拍一方，另一方在镜头后\n \n ✅ 正确（特写只需焦点）：\n \"脸部特写：李四眉头紧锁\" ← 特写不需要交代其他人\n \n ❌ 错误（中景却丢失可见角色）：\n \"中景：李四说话\" ← 中景应该能看到对方，为什么没写？\n\n4. **连续性检查**（生成每个分镜前自检）：\n □ 当前镜头类型/角度能拍摄到哪些角色？\n □ 能拍到的角色是否都有描述？\n □ 拍不到的角色（特写、反打等情况）可以省略\n\n【输出格式】\n\n只返回JSON数组，不得输出markdown代码块标记、注释或解释。\n\n示例（重点展示镜头连续性）：\n\n原文：\"张三走进办公室，看着正在工作的李四和王五说：开会了。李四抬头点了点头，王五放下手中的笔站起身。\"\n\n[\n {\n \"panel_number\": 1,\n \"description\": \"中景：张三推开办公室门走进来，画面深处李四坐在左侧工位低头工作，王五坐在右侧工位写字\",\n \"characters\": [\n {\"name\": \"张三\", \"appearance\": \"初始形象\"},\n {\"name\": \"李四\", \"appearance\": \"初始形象\"},\n {\"name\": \"王五\", \"appearance\": \"初始形象\"}\n ],\n \"location\": \"办公室\",\n \"scene_type\": \"daily\",\n \"source_text\": \"张三走进办公室，看着正在工作的李四和王五\"\n },\n {\n \"panel_number\": 2,\n \"description\": \"近景：张三站在门口开口说话，对面李四和王五抬起头望向他\",\n \"characters\": [\n {\"name\": \"张三\", \"appearance\": \"初始形象\"},\n {\"name\": \"李四\", \"appearance\": \"初始形象\"},\n {\"name\": \"王五\", \"appearance\": \"初始形象\"}\n ],\n \"location\": \"办公室\",\n \"scene_type\": \"daily\",\n \"source_text\": \"说：开会了\"\n },\n {\n \"panel_number\": 3,\n \"description\": \"近景：李四坐在工位上抬头点了点头，旁边王五正在放下手中的笔，背景中张三站在门口等待\",\n \"characters\": [\n {\"name\": \"李四\", \"appearance\": \"初始形象\"},\n {\"name\": \"王五\", \"appearance\": \"初始形象\"},\n {\"name\": \"张三\", \"appearance\": \"初始形象\"}\n ],\n \"location\": \"办公室\",\n \"scene_type\": \"daily\",\n \"source_text\": \"李四抬头点了点头，王五放下手中的笔\"\n },\n {\n \"panel_number\": 4,\n \"description\": \"中景：王五从座位上站起身，左侧李四也准备起身，张三在门口向外走去\",\n \"characters\": [\n {\"name\": \"王五\", \"appearance\": \"初始形象\"},\n {\"name\": \"李四\", \"appearance\": \"初始形象\"},\n {\"name\": \"张三\", \"appearance\": \"初始形象\"}\n ],\n \"location\": \"办公室\",\n \"scene_type\": \"daily\",\n \"source_text\": \"王五站起身\"\n }\n]\n\n注意示例中的镜头连续性技巧：\n- 每个镜头都交代了三个角色的位置\n- 镜头焦点变化时（如镜头3焦点是李四王五），仍用「背景中张三」保持连续性\n- 角色移动（如镜头4张三向外走）有明确动作交代\n\n【输入数据】\n\n角色资产库：{characters_lib_name}\n场景资产库：{locations_lib_name}\n\n角色介绍（⭐用于理解\"我\"和称呼对应的角色）：\n{characters_introduction}\n\n角色形象列表（供选择appearance）：\n{characters_appearance_list}\n\n角色完整描述（供参考）：\n{characters_full_description}\n\nClip信息：\n{clip_json}\n\n内容输入（剧本格式JSON或原文片段）：\n{clip_content}\n\n【严格要求】\n1. 必须输出所需数量的有效分镜，禁止空分镜\n2. 角色和场景名字必须从资产库选择\n3. characters 必须是对象数组：[{name: \"角色名\", appearance: \"形象名\"}]\n4. 只返回JSON数组，不得有其他文字\n5. ⚠️ source_text 必填，不得为空或null\n6. 空间关系必须清晰（朝向、阻挡、位置）\n7. 镜头连续性：前后镜头要有动作承接\n8. 禁止身份称呼：必须使用资产库中的具体名字\n9. 禁止主观情绪词：只描述可视化动作和状态\n10. 禁止长句单镜头：包含逗号分隔多个动作/对话的长句必须拆分\n11. 对话必须拆分：每段对话至少 2 个镜头（说话者 + 听者反应）\n12. ⚠️ 镜头合理性：只描述当前镜头**实际能拍摄到**的角色，特写/反打等拍不到的可省略\n13. ⚠️ JSON安全：原文中的所有引号（\"\"''「」等）在 JSON 字符串值中必须统一替换为「」，严禁出现未转义的英文双引号 \"\n"
},
{
"path": "lib/prompts/novel-promotion/character_create.en.txt",
"content": "You are a professional character prompt designer.\nGenerate one image-ready character appearance prompt from the user's request.\n\nUser request:\n{user_input}\n\nRequirements:\n1. Output one complete English appearance prompt.\n2. Include age range, facial traits, hairstyle, body build, outfit, shoes, and accessories.\n3. Keep it visual and concrete, no story narration.\n4. Do not include expression, action, background, or camera language.\n5. Do not mention skin color, eye color, or lip color.\n6. Keep it concise and production-ready.\n\nOutput format:\nReturn JSON only. ⚠️ JSON SAFETY: All quotation marks in text MUST be converted to corner brackets「」in JSON string values. NEVER use raw ASCII double quotes \" inside string values:\n{\n \"prompt\": \"character appearance prompt\"\n}\n"
},
{
"path": "lib/prompts/novel-promotion/character_create.zh.txt",
"content": "请按照以下提示词规则执行用户的生成人物需求\n【人物生成要求（用于出图，中文描述）】\n\n1. 生成1条详细的中文外貌描述，供AI图片生成使用\n\n2. 描述要求突出角色特色，有细节质感：\n - 性别、年龄范围（写具体年龄区间，如\"约二十五岁\"、\"四十至四十五岁\"、\"五十岁左右\"）\n - 面部：脸型、五官特征（如高挺鼻梁、深邃眼窝、薄唇等具体特征）\n - 眼睛：形状、大小（禁止描写眼睛颜色）\n - 头发：颜色、长度、发型、发质（如蓬松卷发、挑染银灰、发尾微卷等）\n - 体型：身高感、体态、肩宽、腰线等\n - 皮肤：只描述质感和独特标记（如光滑/粗糙、雀斑、胎记、疤痕、纹身等），禁止描述肤色\n - 服装：款式、材质、配色、细节（如机车皮夹克、破洞牛仔裤、金属拉链、刺绣图案等）\n - 鞋子：款式、颜色、材质（如黑色马丁靴、白色帆布鞋、棕色皮质牛津鞋、红底高跟鞋、绣花布鞋等）\n - 配饰：耳钉、项链、手表、戒指等突出个性的配饰\n\n3. 【角色类型判断】\n - 如果用户描述的是非人类角色（动物、神话生物、知名IP形象等），不受上述人类外貌模板限制\n - 描述开头必须以角色名或物种名开始\n - 根据实际形态自由描述，保持核心辨识特征\n\n 示例：\n - 输入\"孙悟空\" → 输出：\"孙悟空，金毛覆身，头戴紧箍咒，身穿虎皮战裙，手持金箍棒...\"\n - 输入\"一只蜗牛\" → 输出：\"蜗牛，背负螺旋形硬壳，两只触角细长探出...\"\n - 输入\"皮卡丘\" → 输出：\"皮卡丘，黄色圆润身体，红色脸颊，闪电形尾巴，尖耳带黑色耳尖...\"\n\n4. 描述规范：\n - 禁止写表情、姿态、动作\n - 禁止写背景/环境/道具\n - 不得加入情绪形容词与故事性句子\n - 【禁止身体颜色描述】禁止描写任何身体部位的颜色，AI会过度放大颜色描述导致效果失真：\n ❌ 皮肤颜色（如偏黄、白皙、小麦色、古铜色、黝黑等）\n ❌ 唇色（如红润、粉色、苍白等）\n ❌ 眼睛颜色（如黑色、棕色、蓝色瞳孔等）\n ❌ 脸色（如红润、苍白、蜡黄等）\n ✅ 可以描写：皮肤质感（光滑/粗糙）、独特标记（雀斑/疤痕/纹身）、头发颜色、服装颜色\n - 如原文对外貌有描述，以原文为最优先（但颜色描述仍需过滤）\n - 使用中文输出，长度 80-150 字\n - 不包含艺术风格、画风、光影效果描述（系统自动添加）\n - 【年代一致性】根据故事背景判断年代（古代/近代/现代/未来等），人物的服装、发型、配饰必须符合该年代特征\n - 【禁止不确定描述】禁止使用\"或\"、\"可能\"、\"也许\"、\"大概\"等不确定词汇，每个外貌特征必须明确具体\n ❌ 错误：\"戴着无框或金框眼镜\"、\"身高可能一米七左右\"\n ✅ 正确：\"戴着金色细框眼镜\"、\"身材高挑约一米七五\"\n - 【禁止抽象气质描述】禁止描述无法视觉化的抽象气质、氛围、神态、感受\n ❌ 错误：\"举手投足间透着富贵气\"、\"气场强大\"、\"成熟稳重的气息\"\n ✅ 正确：只描述可直接看到的外貌特征\n - 【鞋子必填】每个完整人物描述必须包含鞋子描述，不可遗漏\n\n\n你的目标是根据用户发送你的需求以及上述规则生成一个人物提示词\n以下是用户的生成指令:{user_input}\n\n发送json格式给我，只返回以下json格式，禁止返回一切除json以外的多余内容，注释，文字等等，只返回无任何markdown标识符的纯净json格式。⚠️ 所有引号（\"\"''等）在 JSON 字符串值中必须统一替换为「」，严禁出现未转义的英文双引号 \"。json格式如下\n{\n \"prompt\":\"xxxxx\"\n}"
},
{
"path": "lib/prompts/novel-promotion/character_description_update.en.txt",
"content": "You are a character appearance prompt editor.\nUpdate the original character description according to the user's edit instruction.\n\nOriginal description:\n{original_description}\n\nUser instruction:\n{modify_instruction}\n\nReference image context (may be empty):\n{image_context}\n\nRules:\n1. Keep unchanged traits unless user explicitly asks to change them.\n2. Merge requested changes into a single complete prompt.\n3. Output in English only.\n4. No expression, action, background, or props.\n5. No skin color, eye color, or lip color.\n\nOutput format:\nReturn JSON only. ⚠️ JSON SAFETY: All quotation marks MUST be converted to corner brackets「」in JSON string values:\n{\n \"prompt\": \"updated full character description\"\n}\n"
},
{
"path": "lib/prompts/novel-promotion/character_description_update.zh.txt",
"content": "你是一个专业的角色形象描述更新专家。\n\n【任务】\n根据用户对角色图片的修改，更新角色的形象描述词。\n\n【原始角色描述】\n{original_description}\n\n【用户修改指令】\n{modify_instruction}\n\n{image_context}\n\n【更新规则】\n1. 仔细理解用户的修改指令，找出需要修改的具体特征\n2. 如果有参考图片，请识别参考图片中的关键视觉特征（如服装款式、颜色、材质、配饰等）\n3. 将修改内容准确融入原始描述中，替换或补充相关部分\n4. 保持描述的流畅性和一致性\n5. 保留未被修改的原有特征\n6. 遵循以下描述规范：\n - 禁止写表情、姿态、动作\n - 禁止写背景/环境/道具\n - 禁止描写身体部位颜色（皮肤色、唇色、眼睛颜色等）\n - 使用中文输出，长度 80-150 字\n\n【输出格式】\n只返回JSON格式，禁止返回任何其他内容。⚠️ 所有引号（\"\"''等）在 JSON 字符串值中必须替换为「」，严禁出现未转义的英文双引号 \"：\n{\n \"prompt\": \"更新后的完整角色描述\"\n}\n"
},
{
"path": "lib/prompts/novel-promotion/character_modify.en.txt",
"content": "You are a professional character prompt modifier.\nModify an existing character description based on user instruction.\n\nCurrent description:\n{character_input}\n\nUser instruction:\n{user_input}\n\nRules:\n1. Keep identity consistency.\n2. Apply only requested edits, keep other valid details.\n3. Return one complete rewritten prompt, not partial fragments.\n4. Output in English only.\n5. Do not describe expression, action, background, scene, or props.\n6. Do not mention skin color, eye color, or lip color.\n\nOutput format:\nReturn JSON only. ⚠️ JSON SAFETY: All quotation marks MUST be converted to corner brackets「」in JSON string values:\n{\n \"prompt\": \"modified character description\"\n}\n"
},
{
"path": "lib/prompts/novel-promotion/character_modify.zh.txt",
"content": "请按照以下规则执行用户的人物生成提示词修改需求\n1:人物规则按照以下规则修改\n【人物生成要求（用于出图，中文描述）】\n\n1. 生成1条详细的中文外貌描述，供AI图片生成使用\n\n2. 描述要求突出角色特色，有细节质感：\n - 性别、年龄范围（写具体年龄区间，如\"约二十五岁\"、\"四十至四十五岁\"、\"五十岁左右\"）\n - 面部：脸型、五官特征（如高挺鼻梁、深邃眼窝、薄唇等具体特征）\n - 眼睛：形状、大小（禁止描写眼睛颜色）\n - 头发：颜色、长度、发型、发质（如蓬松卷发、挑染银灰、发尾微卷等）\n - 体型：身高感、体态、肩宽、腰线等\n - 皮肤：只描述质感和独特标记（如光滑/粗糙、雀斑、胎记、疤痕、纹身等），禁止描述肤色\n - 服装：款式、材质、配色、细节（如机车皮夹克、破洞牛仔裤、金属拉链、刺绣图案等）\n - 鞋子：款式、颜色、材质（如黑色马丁靴、白色帆布鞋、棕色皮质牛津鞋、红底高跟鞋、绣花布鞋等）\n - 配饰：耳钉、项链、手表、戒指等突出个性的配饰\n\n3. 描述规范：\n - 禁止写表情、姿态、动作\n - 禁止写背景/环境/道具\n - 不得加入情绪形容词与故事性句子\n - 【禁止身体颜色描述】禁止描写任何身体部位的颜色，AI会过度放大颜色描述导致效果失真：\n ❌ 皮肤颜色（如偏黄、白皙、小麦色、古铜色、黝黑等）\n ❌ 唇色（如红润、粉色、苍白等）\n ❌ 眼睛颜色（如黑色、棕色、蓝色瞳孔、琥珀色等）\n ❌ 脸色（如红润、苍白、蜡黄等）\n ✅ 可以描写：皮肤质感（光滑/粗糙）、独特标记（雀斑/疤痕/纹身）、头发颜色、服装颜色\n - 如原文对外貌有描述，以原文为最优先（但颜色描述仍需过滤）\n - 使用中文输出，长度 80-150 字\n - 不包含艺术风格、画风、光影效果描述（系统自动添加）\n - 【年代一致性】根据故事背景判断年代（古代/近代/现代/未来等），人物的服装、发型、配饰必须符合该年代特征\n - 【禁止不确定描述】禁止使用\"或\"、\"可能\"、\"也许\"、\"大概\"等不确定词汇，每个外貌特征必须明确具体\n ❌ 错误：\"戴着无框或金框眼镜\"、\"身高可能一米七左右\"\n ✅ 正确：\"戴着金色细框眼镜\"、\"身材高挑约一米七五\"\n - 【禁止抽象气质描述】禁止描述无法视觉化的抽象气质、氛围、神态、感受\n ❌ 错误：\"举手投足间透着富贵气\"、\"气场强大\"、\"成熟稳重的气息\"\n ✅ 正确：只描述可直接看到的外貌特征\n - 【鞋子必填】每个完整人物描述必须包含鞋子描述，不可遗漏\n - 【非人类角色处理】如果当前角色是非人类（动物、神话生物、知名形象等）：\n * 不受人类外貌模板限制，根据实际形态描述\n * 描述开头保持角色名或物种名\n * 示例：修改\"孙悟空\"的服装 → \"孙悟空，金毛蓬松，换上白色僧袍，头戴紧箍咒...\"\n\n2:你的目标是根据用户发送你的需求将人物修改为符合用户提示词的样子\n\n以下是原本的人物生成提示词:{character_input}\n以下是用户的修改指令:{user_input}\n\n修改后发送json格式给我，只返回以下json格式，禁止返回一切除json以外的多余内容，注释，文字等等，只返回无任何markdown标识符的纯净json格式。⚠️ 所有引号（\"\"''等）在 JSON 字符串值中必须统一替换为「」，严禁出现未转义的英文双引号 \"。json格式如下\n{\n \"prompt\":\"xxxxx\"\n}"
},
{
"path": "lib/prompts/novel-promotion/character_regenerate.en.txt",
"content": "You are a character appearance regenerator.\nGenerate 3 new character appearance variants based on story context.\n\nCharacter name:\n{character_name}\n\nAppearance type / reason:\n{change_reason}\n\nCurrent descriptions (reference only, do not copy directly):\n{current_descriptions}\n\nStory context:\n{novel_text}\n\nRequirements:\n1. Produce 3 clearly different variants while preserving core identity.\n2. Each variant must be a full standalone appearance description.\n3. Output in English.\n4. Do not include expression, action, background, or story narration.\n5. Do not include skin color, eye color, or lip color.\n6. Keep each description concise and image-generation friendly.\n\nOutput format:\nReturn JSON only. ⚠️ JSON SAFETY: All quotation marks MUST be converted to corner brackets「」in JSON string values:\n{\n \"descriptions\": [\n \"variant 1\",\n \"variant 2\",\n \"variant 3\"\n ]\n}\n"
},
{
"path": "lib/prompts/novel-promotion/character_regenerate.zh.txt",
"content": "你是\"角色形象重塑师\"。请根据小说原文，为指定角色重新生成 3 条全新的外貌描述。\n\n【角色信息】\n- 角色名：{character_name}\n- 形象类型：{change_reason}\n- 当前描述（参考，需要生成不同的）：\n{current_descriptions}\n\n【生成要求】\n1. 根据小说原文对该角色的描写，生成 3 条各有特色、互不相同的外貌描述\n2. 必须与当前描述有明显差异（换一种风格/配色/细节），但保持角色核心特征\n3. ⚠️ 【重要】每条描述都必须是【完整的人物描述】，包含所有基础特征（面部、眼睛、体型等）+ 服装/状态，禁止只写变化部分\n4. 【非人类角色处理】\n - 如果角色是非人类（动物、神话生物、知名形象等），不受人类模板限制\n - 每条描述开头必须以角色名或物种名开始\n - 根据实际形态自由描述外观特征\n5. 描述内容：\n - 性别、年龄段（不写具体数字）\n - 面部：脸型、五官特征（如高挺鼻梁、深邃眼窝、薄唇等）\n - 眼睛：形状、大小（禁止描写眼睛颜色）\n - 头发：颜色、长度、发型、发质（如蓬松卷发、挑染银灰等）\n - 体型：身高感、体态、肩宽、腰线等\n - 皮肤：只描述质感和独特标记（如光滑/粗糙、雀斑、疤痕、纹身等），禁止描述肤色\n - 服装：款式、材质、配色、细节（如机车皮夹克、金属拉链等）\n - 鞋子：款式、颜色、材质（如黑色马丁靴、白色帆布鞋、棕色皮质牛津鞋、红底高跟鞋、绣花布鞋等）\n - 配饰：耳钉、项链、手表、戒指等突出个性的配饰\n\n4. 描述规范：\n - 禁止写表情、姿态、动作\n - 禁止写背景/环境/道具\n - 不得加入情绪形容词与故事性句子\n - 【禁止身体颜色描述】禁止描写任何身体部位的颜色，AI会过度放大颜色描述导致效果失真：\n ❌ 皮肤颜色（如偏黄、白皙、小麦色、古铜色、黝黑等）\n ❌ 唇色（如红润、粉色、苍白等）\n ❌ 眼睛颜色（如黑色、棕色、蓝色瞳孔、琥珀色等）\n ❌ 脸色（如红润、苍白、蜡黄等）\n ✅ 可以描写：皮肤质感（光滑/粗糙）、独特标记（雀斑/疤痕/纹身）、头发颜色、服装颜色\n - 如原文对外貌有描述，以原文为最优先（但颜色描述仍需过滤）\n - 使用中文输出，长度 80-150 字\n - 不包含艺术风格、画风、光影效果描述（系统自动添加）\n - 【年代一致性】根据故事背景判断年代，服装、发型、配饰必须符合该年代特征\n\n【输出格式】只返回以下 JSON，不要任何其他内容。⚠️ 所有引号（\"\"''等）在 JSON 字符串值中必须替换为「」，严禁出现未转义的英文双引号 \"。\n{\n \"descriptions\": [\n \"新描述1（80-150字）\",\n \"新描述2（80-150字）\",\n \"新描述3（80-150字）\"\n ]\n}\n\n【小说原文】\n{novel_text}\n\n\n\n\n\n\n\n"
},
{
"path": "lib/prompts/novel-promotion/episode_split.en.txt",
"content": "You are a long-text episode splitter.\nAnalyze the full text and split it into balanced episodes.\n\nInput text:\n{CONTENT}\n\nCore rules:\n1. Keep episode lengths as balanced as possible.\n2. Prefer natural breakpoints (chapter boundary, scene shift, time jump).\n3. Keep chronology and full coverage (no overlap, no missing text).\n4. Provide reliable startMarker and endMarker copied from source text.\n5. Return JSON only.\n6. ⚠️ JSON SAFETY: All quotation marks in text (\"\"''「」 etc.) MUST be converted to corner brackets「」in JSON string values. NEVER use raw ASCII double quotes \" inside string values.\n\nOutput format:\n{\n \"analysis\": {\n \"totalWords\": 0,\n \"episodeCount\": 0,\n \"targetWordsPerEpisode\": 0,\n \"allowedRange\": \"0-0\"\n },\n \"episodes\": [\n {\n \"number\": 1,\n \"title\": \"Episode title\",\n \"summary\": \"Short summary\",\n \"estimatedWords\": 0,\n \"startMarker\": \"exact start marker\",\n \"endMarker\": \"exact end marker\",\n \"startIndex\": 0,\n \"endIndex\": 0\n }\n ],\n \"validation\": {\n \"maxWords\": 0,\n \"minWords\": 0,\n \"variance\": 0,\n \"isBalanced\": true\n }\n}\n"
},
{
"path": "lib/prompts/novel-promotion/episode_split.zh.txt",
"content": "你是一个专业的内容分析助手。请分析以下文本，将其智能分割为多个剧集。\n\n## ⚠️ 核心规则：字数必须均衡（最重要）\n\n**所有剧集的字数必须尽可能均衡！偏差不得超过 ±20%**\n\n### 📊 第一步：精确计算（必须执行）\n\n1. 统计总字数：count_total_characters(文本)\n2. 计算目标集数：total ÷ 650 = N 集（四舍五入）\n3. 计算每集目标字数：target = total ÷ N\n4. 确定允许范围：[target × 0.8, target × 1.2]\n\n**示例**：\n- 总字数 6500 字 → 10 集 → 每集目标 650 字 → 范围 520-780 字\n- 总字数 13000 字 → 20 集 → 每集目标 650 字 → 范围 520-780 字\n\n### 📐 第二步：均衡分割（必须执行）\n\n❌ **绝对禁止**：\n- 任何一集超过 target × 1.3（如目标650字，禁止超过845字）\n- 任何一集少于 target × 0.6（如目标650字，禁止少于390字）\n- 前几集很长、后几集很短（或反过来）\n\n✅ **必须保证**：\n- 所有集的字数在目标值 ±20% 范围内\n- 最长集与最短集的差距不超过 300 字\n- 字数分布均匀，不能头重脚轻或头轻脚重\n\n### 🎬 第三步：寻找分割点\n\n1. **优先识别自然断点**：\n - 章节标记：「第X集」「Chapter X」「Episode X」\n - 场景编号：`X-Y【场景】` 中的 X 变化（1-x → 2-x 是新集）\n - 时间跳跃：「第二天」「三个月后」\n\n2. **在自然断点附近微调**：\n - 如果自然断点导致字数不均，可在附近段落边界调整\n - 优先在对话结束、场景转换处分割\n - 宁可牺牲一点叙事连贯性，也要保证字数均衡\n\n## 输入文本\n\n{{CONTENT}}\n\n## 📝 输出格式\n\n```json\n{\n \"analysis\": {\n \"totalWords\": 6500,\n \"episodeCount\": 10,\n \"targetWordsPerEpisode\": 650,\n \"allowedRange\": \"520-780\"\n },\n \"episodes\": [\n {\n \"number\": 1,\n \"title\": \"剧集标题（4-8字）\",\n \"summary\": \"50字以内的剧情简介\",\n \"estimatedWords\": 650,\n \"startMarker\": \"该集开头的前20个字符（精确复制原文）\",\n \"endMarker\": \"该集结尾的后20个字符（精确复制原文）\"\n }\n ],\n \"validation\": {\n \"maxWords\": 720,\n \"minWords\": 590,\n \"variance\": 130,\n \"isBalanced\": true\n }\n}\n```\n\n## ⚠️ 最终验证清单（输出前必须检查）\n\n在输出之前，你必须验证以下条件：\n\n1. ☐ episodes.length ≈ totalWords ÷ 650（误差 ±1 集）\n2. ☐ 所有 estimatedWords 都在 allowedRange 范围内\n3. ☐ maxWords - minWords ≤ 300 字\n4. ☐ 没有任何一集超过 850 字\n5. ☐ 没有任何一集少于 400 字\n6. ☐ 上一集 endMarker 紧邻下一集 startMarker，无内容遗漏\n7. ☐ endMarker 不包含下一集的任何内容\n\n**如果验证失败，必须重新调整分割点直到通过！**\n\n⚠️ JSON安全：所有引号（\"\"''等）在 JSON 字符串值中必须统一替换为「」，严禁出现未转义的英文双引号 \"。只返回JSON，禁止markdown代码块标记。\n\n## 🔧 场景编号说明\n\n- `X-Y【场景描述】` 格式中，X = 集数，Y = 场景序号\n- 1-1, 1-2, 1-3 都属于第 1 集\n- 2-1 开始第 2 集\n- 分集在 X 变化时进行\n"
},
{
"path": "lib/prompts/novel-promotion/image_prompt_modify.en.txt",
"content": "You are a prompt refinement expert for storyboard image/video generation.\n\nCurrent image prompt:\n{prompt_input}\n\nCurrent video prompt:\n{video_prompt_input}\n\nUser instruction:\n{user_input}\n\nRequirements:\n1. Return an updated image prompt and video prompt.\n2. Keep subject identity and scene continuity unless user asks to change.\n3. Write in concise English.\n4. Image prompt should focus on visual composition.\n5. Video prompt should focus on motion/performance/camera behavior.\n\nOutput format:\nReturn JSON only. ⚠️ JSON SAFETY: All quotation marks MUST be converted to corner brackets「」in JSON string values:\n{\n \"image_prompt\": \"updated image prompt\",\n \"video_prompt\": \"updated video prompt\"\n}\n"
},
{
"path": "lib/prompts/novel-promotion/image_prompt_modify.zh.txt",
"content": "请按照以下提示词规则执行用户的ai生图以及视频运动提示词场景需求\n1. 详细描述镜头角度和地点、动作、人物、环境，也就是谁、和谁、在哪里、干了什么，如果是空镜标题等，那么也要描述出原文想要表达的东西，例如作者名字，或者和原文有关的内容，提示词务必详细完整\n\n2. **场景描述使用规则（重要）：**\n - locations字段：只写场景名字（如\"病房_白天\"）\n - image_prompt字段：直接使用场景名字（如\"病房_白天\"），可以结合具体位置描述（如\"病房_白天的窗边\"）\n - **禁止添加光线、色调、天气描述**：场景资产已包含完整的光线和色调信息，提示词中不要添加\"阳光\"、\"光线\"、\"明亮\"、\"昏暗\"、\"温暖色调\"、\"冷色调\"、\"天气\"等描述\n - **人物所在的位置和互动的地方要和场景提示词中已有的物品有关系，不要出现人物和场景没有的关系进行互动**\n\n3. **人物描述使用规则：**\n - characters字段：只写人物名字（如\"Victor\"）\n - image_prompt字段：直接使用人物名字（如\"Victor\"），可以结合动作和情绪描述（如\"Victor站在门口，表情严肃\"）\n - 写图片生成提示词的时候务必要写明地点+人物名字+情绪+动作\n - 我们这个是有声书第一视角，第一视角默认就是主角的视角，也就是在pov中的人物决策，如无其他意外那么这个就是主角，按照这个名字来进行主视角生成\n\n4. 我们的场景限制只对有效目前人物处在实际发生地点的有效！例如有回忆、旁白等情况，我们不应该完全限制在固定场景之内，而是可以插入多样性的画面\n\n5. 使用中文输出提示词，提示词编写规范按照用简洁连贯的自然语言写明：主体 + 行为 + 环境 + 空间关系\n 例如：Dr.Smith站在病床边，看着躺在床上的Mary，病房_白天\n\n6.涉及到文字内容的全部使用原文srt的语言作为输出，如原文srt是英文，就代表着我们这个面向英文观众，那么如果会画面内容里面出现文字的话（例如出现了医院名字的特写）就要输出英文的具体文字提示词（注意，主要提示词还是中文），如：医院的镜头特写，上面写着hospital\n\n7.使用主体+行为+环境的提示词，确保不要错过主体，除非无主体的空镜，否则一定要交代主体是谁\n\n按照用户的修改指令把提示词修改为用户需要的样子，用户可能会发送一些额外内容让你修改，例如让你把人物修改为另外一个人物，这个时候你会看到人物或场景的具体描述词\n例如用户输入：把张三(黑色长发,穿着西装...)修改为小明(蓝色头发，红色眼睛...)然后就可以利用这些具体描述词来按照以上的生成提示词规则修改。\n\n当前的图片提示词：{prompt_input}\n\n当前的视频提示词：{video_prompt_input}\n\n用户的修改指令：{user_input}\n\n结果发送json格式给我，只返回以下json格式，禁止返回一切除json以外的多余内容，注释，文字等等，只返回无任何markdown标识符的纯净json格式。⚠️ 所有引号（\"\"''等）在 JSON 字符串值中必须统一替换为「」，严禁出现未转义的英文双引号 \"。json格式如下\n{\n \"image_prompt\": \"修改后的图片提示词（静态画面描述）\",\n \"video_prompt\": \"\"\n}\n"
},
{
"path": "lib/prompts/novel-promotion/location_create.en.txt",
"content": "You are a professional environment prompt designer.\nGenerate one scene prompt for image generation.\n\nUser request:\n{user_input}\n\nRules:\n1. Output in English only.\n2. Start with scene name in this format: \"[Scene Name] ...\"\n3. Describe a wide, clear environment with spatial layout and key objects.\n4. Mention lighting direction and atmosphere.\n5. No protagonist actions or dialogue.\n6. If crowd is implied by context, use generic crowd terms only (guests, pedestrians, audience).\n\nOutput format:\nReturn JSON only. ⚠️ JSON SAFETY: All quotation marks MUST be converted to corner brackets「」in JSON string values:\n{\n \"prompt\": \"[Scene Name] environment description\"\n}\n"
},
{
"path": "lib/prompts/novel-promotion/location_create.zh.txt",
"content": "请按照以下提示词规则执行用户的生成场景需求\n\n【场景生成要求（用于出图，中文描述）】\n\n1. 生成1条中文环境描述（60-120字），像真实摄影场景一样描述\n\n2. **开头必须明确写明场景名称**：\n - 描述开头必须以\"【场景名称】\"的形式标注空间属性\n - 示例：「皇宫」殿内铺设着... / 「客厅」窗外阳光透过... / 「卧室」床边放着...\n - 这样AI在生成图片时能明确理解这是什么类型的空间\n\n3. 核心原则：\n - 写真实存在的物体，不要写抽象感受\n - 材质要具体（深棕色实木地板、青灰色石砖墙、做旧铁艺栏杆）\n - 物品要有使用痕迹和生活气息（桌上散落的书籍、墙角堆放的杂物、窗台晒干的植物）\n - 光线要写清楚来源和效果（午后阳光斜照进来在地板上拉出长影、暖黄色壁灯打在墙面上）\n\n4. 禁止：不写主角人物具体动作、不写画风、不写\"温馨\"\"优雅\"等抽象词\n\n5. 【人群处理规则】\n - 如果用户描述的场景暗示有人群（如宴会厅、集市、教室上课等），在描述中加入模糊人群元素\n - 人群描述示例：\"大厅中宾客三两成群\"、\"街道上行人往来\"、\"座位上零散坐着几位观众\"\n - 如果是私密空间或用户明确要求空镜，则不添加人群\n\n以下是用户的生成指令:{user_input}\n\n只返回以下json格式，禁止返回一切除json以外的多余内容。⚠️ 所有引号（\"\"''等）在 JSON 字符串值中必须替换为「」，严禁出现未转义的英文双引号 \"。\n{\n \"prompt\":\"「场景名称」场景描述内容\"\n}\n"
},
{
"path": "lib/prompts/novel-promotion/location_description_update.en.txt",
"content": "You are a scene description editor.\nUpdate the original location description based on user instruction.\n\nLocation name:\n{location_name}\n\nOriginal description:\n{original_description}\n\nUser instruction:\n{modify_instruction}\n\nReference image context (may be empty):\n{image_context}\n\nRules:\n1. Keep unchanged scene elements unless explicitly modified.\n2. Return one complete updated description in English.\n3. Keep scene name at the beginning: \"[{location_name}] ...\"\n4. No protagonist actions or story narration.\n\nOutput format:\nReturn JSON only. ⚠️ JSON SAFETY: All quotation marks MUST be converted to corner brackets「」in JSON string values:\n{\n \"prompt\": \"updated location description\"\n}\n"
},
{
"path": "lib/prompts/novel-promotion/location_description_update.zh.txt",
"content": "你是一个专业的场景描述更新专家。\n\n【任务】\n根据用户对场景图片的修改，更新场景的描述词。\n\n【场景名称】\n{location_name}\n\n【原始场景描述】\n{original_description}\n\n【用户修改指令】\n{modify_instruction}\n\n{image_context}\n\n【更新规则】\n1. **开头必须明确写明场景名称**：\n - 描述开头必须以「{location_name}」的形式标注空间属性\n - 示例：「皇宫」殿内铺设着... / 「客厅」窗外阳光透过...\n - 这样AI在生成图片时能明确理解这是什么类型的空间\n\n2. 仔细理解用户的修改指令，找出需要修改的具体特征\n3. 如果有参考图片，请识别参考图片中的关键视觉特征（如建筑风格、装饰元素、光线氛围、色调等）\n4. 将修改内容准确融入原始描述中，替换或补充相关部分\n5. 保持描述的流畅性和一致性\n6. 保留未被修改的原有特征\n7. 遵循以下描述规范：\n - 只描述场景本身，禁止描述人物\n - 使用中文输出，长度 50-100 字\n\n【输出格式】\n只返回JSON格式，禁止返回任何其他内容。⚠️ 所有引号（\"\"''等）在 JSON 字符串值中必须替换为「」，严禁出现未转义的英文双引号 \"：\n{\n \"prompt\": \"「场景名」更新后的完整场景描述\"\n}\n"
},
{
"path": "lib/prompts/novel-promotion/location_modify.en.txt",
"content": "You are a professional scene prompt modifier.\nModify an existing scene description while preserving scene identity.\n\nLocation name:\n{location_name}\n\nCurrent description:\n{location_input}\n\nUser instruction:\n{user_input}\n\nRules:\n1. Keep core scene identity and function.\n2. Apply requested changes with concrete visual details.\n3. Output in English only.\n4. Start with scene name: \"[{location_name}] ...\"\n5. No protagonist actions, dialogue, or narrative plot.\n\nOutput format:\nReturn JSON only. ⚠️ JSON SAFETY: All quotation marks MUST be converted to corner brackets「」in JSON string values:\n{\n \"prompt\": \"modified location description\"\n}\n"
},
{
"path": "lib/prompts/novel-promotion/location_modify.zh.txt",
"content": "请按照以下提示词规则执行用户的修改场景需求\n\n【场景名称】\n{location_name}\n\n【场景生成要求（用于出图，中文描述）】\n\n1. **开头必须明确写明场景名称**：\n - 描述开头必须以「{location_name}」的形式标注空间属性\n - 示例：「皇宫」殿内铺设着... / 「客厅」窗外阳光透过...\n - 这样AI在生成图片时能明确理解这是什么类型的空间\n\n2. 每个场景生成 1 条中文环境描述，用于AI图片生成\n\n3. 必须包含具体的视觉元素，按以下结构描述：\n\n **空间定位**：明确场景类型\n - 室内：客厅/卧室/厨房/办公室/医院病房/教室等\n - 室外：街道/巷子/公园/山谷/海边/广场等\n \n **主要结构**：描述可见的建筑元素\n - 墙面/地面/天花板的材质和颜色（如\"白色墙面\"/\"木质地板\"/\"水泥地面\"/\"瓷砖地面\"）\n - 门窗位置和类型（如\"左侧有大窗户\"/\"右侧木门\"/\"落地窗\"/\"玻璃门\"）\n - 建筑特征（如\"高天花板\"/\"拱形门\"/\"砖墙\"/\"玻璃幕墙\"）\n \n **家具/道具**：列出场景中的主要物体\n - 家具摆放（如\"中央有沙发\"/\"墙边书架\"/\"床靠窗\"/\"办公桌\"）\n - 特征道具（如\"桌上有台灯\"/\"墙上挂画\"/\"地上地毯\"/\"书架上有书\"）\n - 植物装饰（如\"窗台有盆栽\"/\"角落有绿植\"）\n \n **光线环境**：描述光源和照明效果\n - 自然光：时间段+光线特征\n * 白天：窗外阳光透过窗帘/明亮日光从窗户照入/柔和晨光\n * 夜晚：月光从窗外照入/窗外夜色/窗外灯光\n - 人工光：光源类型+位置\n * 天花板吊灯照明/墙壁壁灯柔和光线/台灯局部照明/落地灯/射灯\n \n **氛围细节**：补充环境特征\n - 整体色调：暖色调/冷色调/中性色/明亮/昏暗\n - 空间感：宽敞/狭窄/纵深感强/开阔/封闭\n - 状态特征：整洁/凌乱/陈旧/现代/简约/复古\n\n4. 描述规范：\n - 使用具体名词，避免抽象形容词\n * ✅ 好：\"木质书桌\"/\"灰色布艺沙发\"/\"白色窗帘\"\n * ❌ 差：\"优雅的家具\"/\"舒适的环境\"/\"温馨的氛围\"\n - 描述固定元素，不写主角人物具体动作、情绪\n - 【人群处理规则】如果用户要求添加人群，或场景本身暗示有人群（如宴会、集市等）：\n * 可以加入模糊人群描述：\\\"人群\\\"、\\\"宾客\\\"、\\\"路人\\\"等\n * 示例：\\\"大厅远处三两宾客交谈\\\"、\\\"街角有行人匆匆走过\\\"\n - 长度控制在 60-100 字\n - 不包含艺术风格描述（如\"美式漫画风\"/\"水彩风\"），风格由系统自动添加\n - 不包含光影效果描述（如\"dramatic lighting\"/\"柔和渐变\"），由风格控制\n\n你的目标是根据用户的修改指令，在原有场景描述的基础上进行修改\n\n当前场景描述：{location_input}\n\n用户的修改指令：{user_input}\n\n发送json格式给我，只返回以下json格式，禁止返回一切除json以外的多余内容，注释，文字等等，只返回无任何markdown标识符的纯净json格式。⚠️ 所有引号（\"\"''等）在 JSON 字符串值中必须统一替换为「」，严禁出现未转义的英文双引号 \"。json格式如下\n{\n \"prompt\":\"「场景名」xxxxx\"\n}\n"
},
{
"path": "lib/prompts/novel-promotion/location_regenerate.en.txt",
"content": "You are a scene variant regenerator.\nGenerate 3 new scene description variants for the same location.\n\nLocation name:\n{location_name}\n\nCurrent descriptions (reference):\n{current_descriptions}\n\nRequirements:\n1. Generate 3 clearly different but same-location variants.\n2. Keep the scene name prefix in each line: \"[{location_name}] ...\"\n3. Output in English only.\n4. Keep environment-only description (no protagonist actions).\n5. Keep each variant concise and image-generation friendly.\n\nOutput format:\nReturn JSON only. ⚠️ JSON SAFETY: All quotation marks MUST be converted to corner brackets「」in JSON string values:\n{\n \"descriptions\": [\n \"[{location_name}] variant 1\",\n \"[{location_name}] variant 2\",\n \"[{location_name}] variant 3\"\n ]\n}\n"
},
{
"path": "lib/prompts/novel-promotion/location_regenerate.zh.txt",
"content": "你是\"场景重塑师\"。请根据当前的场景描述，为指定场景重新生成 3 条全新的场景描述变体。\n\n【场景信息】\n- 场景名：{location_name}\n- 当前描述（作为参考，需要生成不同的变体）：\n{current_descriptions}\n\n【生成要求】\n1. **开头必须明确写明场景名称**：\n - 每条描述开头必须以「{location_name}」的形式标注空间属性\n - 示例：「皇宫」殿内铺设着... / 「客厅」窗外阳光透过...\n - 这样AI在生成图片时能明确理解这是什么类型的空间\n\n2. 根据当前描述的核心元素，生成 3 条各有特色、互不相同的场景描述变体\n3. 必须与当前描述有明显差异（换一种氛围/布局/细节），但保持场景核心特征\n4. 描述内容应包含：\n - 空间结构：整体布局、空间大小、层次感\n - 建筑/地形：建筑风格、地形特征、主要构造物\n - 光影氛围：光源类型、明暗对比、色调倾向\n - 材质细节：地面、墙面、物体的材质质感\n - 环境元素：植物、天气、装饰物等\n - 独特标识：该场景的标志性元素或特殊物件\n\n5. 描述规范：\n - 禁止写主角人物具体动作、剧情\n - 【人群处理规则】如果当前描述中包含人群元素，新描述也应保持人群元素\n * 可以调整人群的位置、密度、状态，但保持有人群存在\n * 人群描述使用模糊词汇：\"人群\"、\"宾客\"、\"路人\"等\n - 使用中文输出，长度 80-150 字\n - 不包含艺术风格、画风描述（系统自动添加）\n - 【年代一致性】根据场景特征判断年代，建筑、装饰、物品必须符合该年代特征\n - 【时间一致性】如场景名包含\"白天/黑夜/黄昏\"等，描述中的光影必须匹配\n\n【输出格式】只返回以下 JSON，不要任何其他内容。⚠️ 所有引号（\"\"''等）在 JSON 字符串值中必须替换为「」，严禁出现未转义的英文双引号 \"。\n{\n \"descriptions\": [\n \"「场景名」新描述1（80-150字）\",\n \"「场景名」新描述2（80-150字）\",\n \"「场景名」新描述3（80-150字）\"\n ]\n}\n"
},
{
"path": "lib/prompts/novel-promotion/screenplay_conversion.en.txt",
"content": "You are a screenplay conversion specialist.\nConvert the clip text into structured screenplay JSON without adding new story facts.\n\nClip ID:\n{clip_id}\n\nClip content:\n{clip_content}\n\nLocation library:\n{locations_lib_name}\n\nCharacter library:\n{characters_lib_name}\n\nCharacter introductions:\n{characters_introduction}\n\nOutput format (JSON object only):\n{\n \"clip_id\": \"{clip_id}\",\n \"original_text\": \"original clip text\",\n \"scenes\": [\n {\n \"scene_number\": 1,\n \"heading\": {\n \"int_ext\": \"INT or EXT\",\n \"location\": \"location name\",\n \"time\": \"morning/day/evening/night\"\n },\n \"description\": \"scene setup\",\n \"characters\": [\"Character A\", \"Character B\"],\n \"content\": [\n {\n \"type\": \"action\",\n \"text\": \"action description\"\n },\n {\n \"type\": \"dialogue\",\n \"character\": \"Character A\",\n \"parenthetical\": \"optional performance cue\",\n \"lines\": \"spoken line\"\n },\n {\n \"type\": \"voiceover\",\n \"character\": \"Narrator or Character\",\n \"text\": \"voiceover content\"\n }\n ]\n }\n ]\n}\n\nRules:\n1. Preserve original story facts; do not invent new events.\n2. Keep scene/content order aligned with source text.\n3. Resolve character aliases to canonical names when possible.\n4. Use the best matching location name from library; if none fits, use source location text.\n5. Return strict JSON only.\n6. ⚠️ JSON SAFETY: All quotation marks in dialogue (\"\"''「」 etc.) MUST be converted to corner brackets「」in JSON string values. NEVER use raw ASCII double quotes \" inside string values—they break JSON structure.\n"
},
{
"path": "lib/prompts/novel-promotion/screenplay_conversion.zh.txt",
"content": "你是专业的编剧和剧本改编师。你的任务是将小说/文学文本转换为标准的影视剧本格式。\n\n⚠️⚠️⚠️【最高优先级原则 - 100%忠实原文】⚠️⚠️⚠️\n\n你的工作是**格式转换**，不是**创作**！你必须100%忠实于原文，严禁任何形式的\"创造性发挥\"：\n\n🚫 绝对禁止：\n- 添加原文中没有的对话、动作、场景描述\n- 扩展或改编原文内容（即使你觉得\"更合理\"或\"更生动\"）\n- 添加原文没有提及的角色反应、表情、心理活动\n- 脑补原文没有描写的环境细节、道具、氛围\n- 添加过渡性描述来\"填充空白\"\n- 用你的理解\"补全\"原文的留白\n\n✅ 你只能做：\n- 将原文已有的内容转换为剧本格式\n- 识别原文明确描述的场景、角色、对话\n- 提取原文已有的动作和环境描述\n\n如果原文内容简短或留白，你的剧本也应该简短，不要试图\"丰富\"它！\n\n【核心职责】\n\n1. 把小说文字转换为剧本格式（场景、对话、动作描述）\n2. 提取场景的时间、地点、环境信息\n3. 识别场景中出现的角色（匹配资产库）\n4. 区分动作、对话、画外音等不同类型\n5. ⭐100%保持原文信息的完整性，不增不减\n\n【剧本格式规范】\n\n标准剧本包含以下元素：\n\n1. **场景头(Scene Heading)**:\n - 格式: 内景/外景 + 地点 + 时间\n - 示例: \"内景 客厅 清晨\" / \"外景 取经路 白天\"\n\n2. **场景描述(Scene Description)**:\n - 简洁描述场景环境、布局、关键道具\n - 不需要过度细节，只需要建立基本视觉印象\n\n3. **动作描述(Action)**:\n - 描述角色的动作、表情、行为\n - 连续的段落形式，不要拆分成碎片\n\n4. **对话(Dialogue)**:\n - 角色名字\n - 副文本(parenthetical): 括号内的表演指导\n - 台词内容\n\n5. **画外音(Voiceover)**:\n - 旁白、独白、回忆、读信等不在画面中的声音\n - 标记角色（如果是特定角色的独白）\n\n【转换规则】\n\n## 1. 场景识别规则\n\n- 分析原文，识别场景边界（地点变化、时间跨越）\n- 每个场景必须包含：\n * 内景(INT)或外景(EXT)\n * 地点名称：\n - 优先从场景资产库选择完全一致的名称\n - ⚠️【重要】如果资产库中没有匹配的场景，直接使用原文中的场景名称，不要强行匹配错误的资产库场景\n - 宁可输出资产库中不存在的场景名，也不要用错误的场景名（后续会自动创建缺失的资产）\n * 时间段（清晨/上午/正午/下午/黄昏/夜晚/深夜）\n\n## 2. 内容类型识别\n\n必须准确区分以下类型：\n\n**action** - 动作描述\n- 角色的动作、表情、行为\n- 场景变化、环境细节\n- 示例: \"孙悟空举起金箍棒，朝六耳猕猴砸去。\"\n\n**dialogue** - 对话\n- 画面中角色的说话\n- 必须包含：character（角色名）、lines（台词）\n- 可选包含：parenthetical（副文本，如\"愤怒地\"\"小声\"）\n- 示例: \n 角色: 孙悟空\n 副文本: 愤怒地\n 台词: \"一个冒牌货，也敢拦你孙爷爷的路！\"\n\n**voiceover** - 画外音\n- 旁白、独白、回忆中的声音、心理活动\n- 不在画面中出现的声音\n- 示例: \"原来孙悟空真的死在了取经路上。\" （旁白）\n- 示例: 二郎神独白：\"猴子死了，我却没有出手...\" （特定角色的画外音）\n\n## 3. 角色识别规则\n\n- 优先从角色资产库中选择完全一致的名字\n- ⚠️【重要】如果资产库中没有匹配的角色，直接使用原文中的角色名称，不要强行匹配错误的资产库角色\n- 宁可输出资产库中不存在的角色名，也不要用错误的角色名（后续会自动创建缺失的资产）\n- 不要使用简称：用\"六耳猕猴\"而非\"六耳\"\n- 不要使用代词：用具体名字替代\"他\"\"她\"\n- characters数组只包含画面中出现的角色，不包括画外音角色\n\n## 4. 副文本(Parenthetical)提取规则\n\n从原文中识别并提取表演指导：\n- \"XX愤怒地说\" → parenthetical: \"愤怒地\"\n- \"XX小声嘀咕\" → parenthetical: \"小声\"\n- \"XX（转身）说\" → parenthetical: \"转身\"\n- \"XX边走边说\" → parenthetical: \"边走边\"\n\n## 5. 动作连续性规则\n\n- 动作描述应该是连续的段落，不要过度拆分\n- 多个连续动作可以合并在一个action中\n- 示例: ❌ 不好: \"孙悟空站起来。\" + \"孙悟空走向门口。\" + \"孙悟空打开门。\"\n- 示例: ✅ 好: \"孙悟空站起来，走向门口，打开门。\"\n\n【输出格式】\n\n只返回JSON对象，不得有markdown代码块标记、注释或解释。\n\n{\n \"clip_id\": \"{clip_id}\",\n \"original_text\": \"原文内容\",\n \n \"scenes\": [\n {\n \"scene_number\": 1,\n \"heading\": {\n \"int_ext\": \"INT或EXT\",\n \"location\": \"场景名称（必须从资产库选择）\",\n \"time\": \"时间段\"\n },\n \"description\": \"场景环境描述\",\n \"characters\": [\"角色1\", \"角色2\"],\n \n \"content\": [\n {\n \"type\": \"action\",\n \"text\": \"动作描述文本\"\n },\n {\n \"type\": \"dialogue\",\n \"character\": \"角色名\",\n \"parenthetical\": \"副文本\",\n \"lines\": \"台词内容\"\n },\n {\n \"type\": \"voiceover\",\n \"character\": \"角色名或旁白\",\n \"text\": \"画外音内容\"\n }\n ]\n }\n ]\n}\n\n【场景描述撰写规则】\n\ndescription字段应该简洁但信息丰富，包含：\n1. 环境类型（室内/室外/特殊环境）\n2. 主要布局（如\"狭长的走廊\"\"开阔的广场\"）\n3. 关键道具或环境元素（如\"右侧落地窗\"\"远处山脉\"）\n4. 氛围提示（如\"荒凉\"\"温馨\"\"阴森\"）\n\n示例：\n- \"简约客厅。米白色墙面，木质地板。右侧落地窗，左侧入口门。\"\n- \"荒凉的取经路。黄沙漫天，远处是连绵的山脉。\"\n- \"阴暗的洞穴。石壁潮湿，只有微弱的火光。\"\n\n【特殊情况处理】\n\n1. **第一人称叙述**:\n - 如果原文是\"我走进房间\"，需要替换为具体角色名\n - ⭐ 参考【角色介绍】中的说明，找到\"我\"对应的角色\n - 如果角色介绍中说明\"我\"对应某角色，则使用该角色名\n - 如果资产库有\"我\"这个角色，则使用\"我\"\n\n2. **称呼映射**:\n - ⭐ 参考【角色介绍】中的称呼说明\n - 如\"老公\"在介绍中说明对应\"林墨\"，则dialogue的character填\"林墨\"\n - 不要被原文的称呼误导，以资产库的名字为准\n\n3. **回忆/闪回场景**:\n - 回忆中的对话/动作 → 正常处理（因为画面中会演）\n - 回忆的旁白叙述 → 使用voiceover类型\n\n4. **多个小场景**:\n - 如果原文包含多个地点变化，拆分成多个scene\n - 每个scene有独立的scene_number\n\n5. **心理活动**:\n - 角色的内心想法 → voiceover类型\n - 标记character为对应角色\n\n【严格要求 - 必须遵守】\n\n⭐⭐⭐ 忠实原文的强制要求 ⭐⭐⭐\n\n1. 🚨【最重要】禁止编造！所有动作、对话、描述必须来自原文，不能添加任何原文中没有的内容\n2. 🚨 如果原文只有一句话，剧本也只能有一句话对应的内容，禁止\"扩写\"\n3. 🚨 如果原文没有描述角色的表情/动作，禁止添加\"XX露出微笑\"、\"XX点了点头\"等内容\n4. 🚨 如果原文没有环境描写，description字段只写能从原文推断的最基本信息\n5. 🚨 禁止添加过渡性动作，如\"XX走了过来\"、\"XX转身离开\"（除非原文明确写了）\n\n格式要求：\n\n1. 只返回JSON对象，不得有markdown标记或注释\n2. location优先从场景资产库选择；如果资产库没有匹配的，使用原文场景名称（宁可缺失也不用错误的）\n3. characters优先从角色资产库选择；如果资产库没有匹配的，使用原文角色名称（宁可缺失也不用错误的）\n4. ⭐ 根据角色介绍中的称呼映射，将原文中的\"我\"、\"老公\"等替换为正确的角色名\n5. ❌ 严禁添加原文中没有的内容（这是最常见的错误！）\n6. content数组保持时间顺序\n7. type只能是: action, dialogue, voiceover\n8. dialogue必须包含character和lines\n9. voiceover如果是特定角色必须包含character\n10. parenthetical是可选的，只在原文有明确表演指导时添加\n11. 输出必须是**严格合法的JSON**：字符串中不能出现原始换行/回车/制表符，必须使用转义字符（\\\\\\\\n、\\\\\\\\r、\\\\\\\\t）\n12. clip_id 必须与输入的 Clip ID 完全一致，严禁输出 \"{clip_id}\" 这种占位符\n13. 建议输出为单行JSON对象（不包含多余空行/解释）\n\n⚠️⚠️⚠️【JSON安全输出 - 最高优先级】⚠️⚠️⚠️\n- 原文中的所有引号（\"\"''「」『』等）在 JSON 字符串值中必须统一替换为日式方括号引号「」\n- ❌ 严禁在 JSON 字符串值中出现英文双引号 \" ！会破坏 JSON 结构！\n- ✅ 正确：\"lines\":\"孙悟空怒道，「一个冒牌货，也敢拦你孙爷爷的路！」\"\n- ❌ 错误：\"lines\":\"孙悟空怒道，\"一个冒牌货，也敢拦你孙爷爷的路！\"\"\n- 如果字符串内确实需要英文双引号，必须用 \\\" 转义\n- 这条规则适用于 lines、text、description、original_text 等所有字符串字段\n\n⚠️ 自检清单（输出前必须确认）：\n- [ ] 我的每一句动作描述都能在原文中找到对应吗？\n- [ ] 我的每一句对话都是原文的原话吗？\n- [ ] 我有没有添加原文没有的\"走过来\"、\"点头\"、\"微笑\"等动作？\n- [ ] 我有没有\"丰富\"原文简短的描写？\n\n【输入数据】\n\nClip原文：\n{clip_content}\n\n场景资产库（优先匹配，无匹配时可使用原文名称）：\n{locations_lib_name}\n\n角色资产库（优先匹配，无匹配时可使用原文名称）：\n{characters_lib_name}\n\n角色介绍（⭐重要：用于理解\"我\"和称呼对应的角色以及角色关系）：\n{characters_introduction}\n\nClip ID:\n{clip_id}\n\n【输出要求】\n\n请将上述原文转换为标准剧本格式，只返回JSON对象。\n再次强调：输出必须是**严格合法的JSON**，不得包含任何额外文本。\n"
},
{
"path": "lib/prompts/novel-promotion/select_location.en.txt",
"content": "You are a location asset extraction specialist.\nExtract locations that need dedicated background assets.\n\nInput text:\n{input}\n\nExisting location library:\n{locations_lib_name}\n\nSelection rules:\n1. Include locations where meaningful story actions happen.\n2. Exclude abstract/metaphorical spaces and one-off passing mentions.\n3. Deduplicate aliases of the same location.\n4. Prefer exact library names when a location already exists.\n\nFor each selected location, generate 3 wide-angle environment descriptions.\nEach description should:\n- start with location name in brackets: \"[Location Name] ...\"\n- describe spatial layout, depth layers, major objects, and lighting direction\n- remain environment-only (no named protagonist actions)\n- use concise, production-ready English\n\nOutput format (JSON only):\n{\n \"locations\": [\n {\n \"name\": \"location_name\",\n \"summary\": \"short usage summary\",\n \"has_crowd\": false,\n \"crowd_description\": \"\",\n \"descriptions\": [\n \"[location_name] description 1\",\n \"[location_name] description 2\",\n \"[location_name] description 3\"\n ]\n }\n ]\n}\n\nStrict constraints:\n1. JSON only.\n2. If no valid location exists, return: {\"locations\":[]}.\n3. ⚠️ JSON SAFETY: All quotation marks in text (\"\"''「」 etc.) MUST be converted to corner brackets「」in JSON string values. NEVER use raw ASCII double quotes \" inside string values.\n "
},
{
"path": "lib/prompts/novel-promotion/select_location.zh.txt",
"content": "你是\"场景资产建立师\"。请基于我提供的文本（可能是小说、剧本、或混合格式），筛选【需要制作画面的场景】，生成用于出图与后续生产的资产 JSON。\n\n【筛选规则 - 精准提取模式】\n\n✅【必须提取的场景】：\n - 剧本场景头部中出现的地点（如\"内景 客厅 白天\"）\n - 角色实际身处、产生互动的具体场所\n - 剧情主线发生的核心地点\n - 多次出现或戏份较重的场景\n - 有明确空间描写、需要制作背景画面的地点\n\n❌【不提取的场景】（严格执行！）：\n - 一次性路过、仅提及但无剧情发生的地点\n - 意境类、比喻类、修辞类描述（如\"从天堂打到地狱\"、\"从天上打到地下\"、\"心灵深处\"、\"记忆长河\"等）\n - 抽象空间或无法具象化的概念（如\"命运交汇点\"、\"时空裂缝\"）\n - 仅作为对话背景提及、没有实际画面需求的地点\n - 纯过渡性场景（如\"穿过走廊\"、\"路过门口\"等一笔带过的移动描述）\n - 回忆/幻想中一闪而过、没有具体剧情的场景\n - 战斗过程中一笔带过的地点（如\"打遍三界\"、\"从山上打到山下\"、\"从天宫打到凡间\"等表示战斗范围的修辞）\n\n📋【判断标准】：\n 问自己：这个场景是否需要单独制作一张背景图？角色是否在此场景有实际戏份？\n 如果只是一句话带过的地点，则不提取。\n 如果是表示\"打斗范围\"的修辞（如从天堂到地狱），则不提取。\n\n🔄【去重规则】：\n - 若场景在库中已存在则跳过，场景库如下：{locations_lib_name}\n - 同一场景不同称呼合并为一个（如\"书房\"和\"张先生的书房\"视为同一场景）\n - 返回的场景名必须与资产库中已有名称完全一致\n\n【场景生成要求 - A: 全景空间版】\n侧重点：宽广完整的空间全貌、整体布局、画面层次\n\n⚠️ 【核心要求】必须生成【宽广的空间全景】，展示场景的完整面貌，而非局部特写！\n- 镜头应该是【广角/远景】视角，能看到整个空间的全貌\n- 展示空间的完整边界（墙壁、地面、天花板/天空）\n- 让观众能够清晰理解这是一个什么样的完整空间\n- 严格按照原文的场景描述来描写，原文描述的场景是最优先级，其他才可以自由发挥\n\n1. **开头必须明确写明场景名称**：\n - 每条描述开头必须以「场景名」的形式标注空间属性\n - 示例：「皇宫」殿内铺设着... / 「客厅」窗外阳光透过... / 「卧室」床边放着...\n - 这样AI在生成图片时能明确理解这是什么类型的空间\n\n2. 每个场景生成 3 条中文环境描述（用于AI图片生成），供用户选择\n\n3. 3条描述要求：\n - 全部符合原文描述的场景特征\n - 可以自由发挥细节，但整体风格保持一致，不要有过大差异\n - 全部使用广角/远景视角，展示完整空间全貌\n - 每条描述开头都必须以「场景名」标注\n\n4. 每条描述都必须包含：\n\n **宽广空间感**（最重要）：\n - 必须是【广角镜头】或【远景视角】，能看到空间的大部分区域\n - 室内场景：能看到2-3面墙壁、地板、部分天花板\n - 室外场景：能看到开阔的视野、远处的地平线或建筑群\n - 强调空间的【开阔感】和【完整性】\n\n **空间定位与规模**：\n - 场景类型（室内/室外/幻想空间）\n - 空间大小感：描述实际的空间尺度（如\"约30平米的客厅\"/\"一眼望不到边的草原\"）\n - 层高/纵深感：能看到的最远距离\n\n **空间层次**（创造画面深度）：\n - 前景：靠近镜头的元素（桌角/门框边缘/植物叶片/栏杆等，部分可见）\n - 中景：主要场景区域（核心物体的完整呈现）\n - 背景：远处可见的元素（窗外景色/远处墙面/天际线/门廊深处）\n\n **物体布局**：\n - 使用明确的位置词：左侧/右侧/中央/角落/靠窗/远处\n - 描述物体之间的空间关系和前后层次\n - 5-8件物体，每件都有位置说明\n\n **光线方向**：光从哪个方向照入，照亮哪些区域\n\n5. 描述规范：\n - 强调位置关系词：前方、远处、左侧、角落、靠近、深处\n - 长度 100-150 字\n \n ⚠️【场景图不能出现任何角色 - 核心规则】：\n \n 场景图的用途：场景图是纯粹的\"背景板\"，主角和重要角色会在后期通过 AI 合成到背景上。\n 因此，场景描述中**绝对不能出现任何有名有姓的角色**。\n \n ❌ 错误示例（包含了角色）：\n - \"两只猴王持棒对峙\" → 错！猴王是角色，不能出现\n - \"张三站在门口迎接\" → 错！张三是角色，不能出现\n - \"孙悟空和六耳猕猴在街上打斗\" → 错！主角不能出现\n \n ✅ 正确示例（纯背景）：\n - \"「古道」广角镜头展现蜿蜒在险峻石林间的黄土古道，前景几株枯松，中景道路宽阔平坦，尘土飞扬，背景是连绵群山。\"\n - \"「宴会厅」大厅远处三两宾客交谈\" → 可以！这是无名背景群众\n - \"「集市」街道上行人往来\" → 可以！这是模糊的路人群众\n \n 📋 什么情况可以写人群？\n - 只有无名的、模糊的背景群众可以出现（如\"宾客\"、\"路人\"、\"行人\"、\"围观群众\"）\n - 这些群众不能有具体描述，只能用模糊词汇\n - 如果场景是私密空间或无人场景，保持空镜即可\n \n - 不包含艺术风格描述，风格由系统自动添加\n\n6. 场景命名规则：中文 \"地点_时间/状态\"\n - 示例：\"客厅_白天\"/\"空间站_夜间\"/\"仙宫_黄昏\"/\"森林_迷雾中\"\n\n7. 剧情中出现的关键元素必须在场景中体现（如椅子、桌子等）\n\n8.如无特殊要求，使用用户输入的语言来进行场景生成，例如输入英文输出偏西方场景，中文则输出偏中国场景，但是原则要按照文字剧本里实际发生的地点为准，\n\n【输出规范（只允许以下 JSON 结构；字段名中文；不得输出任何多余文字）】\n{\n \"locations\": [\n {\n \"name\": \"场景_时间\",\n \"summary\": \"场景简要说明（用途/人物关联，如：张三居住的主卧室、公司高层会议室等）\",\n \"has_crowd\": true/false,\n \"crowd_description\": \"人群类型描述（仅当has_crowd为true时填写，如：宴会宾客、集市人群、学生们等）\",\n \"descriptions\": [\n \"「场景名」场景环境描述1（如has_crowd为true则包含人群元素）\",\n \"「场景名」场景环境描述2\",\n \"「场景名」场景环境描述3\"\n ]\n }\n ]\n}\n\n【严格性】\n- 若无符合条件的场景，locations数组返回 []。\n- 只返回上述 JSON；不得输出markdown代码块标记、如```json注释或解释；不得添加未定义字段。\n- 每条描述必须遵守长度限制（100-150字）；发现超长请自行截断。\n- 禁止在 JSON 字符串值中出现英文双引号 \"。原文中的所有引号（\"\"''等）必须统一替换为「」。如字符串内确实需要英文双引号，必须转义为 \\\"\n\n【原文内容如下】\n{input}\n"
},
{
"path": "lib/prompts/novel-promotion/single_panel_image.en.txt",
"content": "You are a professional storyboard image artist.\nGenerate exactly one high-quality image for one panel.\n\nAbsolute constraints:\n1. No text in the image.\n2. No subtitles, labels, numbers, watermarks, or symbols.\n3. Do not create collage or multi-frame output.\n4. Output exactly one frame.\n\nAspect ratio (must be exact):\n{aspect_ratio}\n\nStoryboard panel data:\n{storyboard_text_json_input}\n\nSource text:\n{source_text}\n\nStyle requirement:\n{style}\n\nExecution rules:\n1. Respect panel composition, character placement, and action logic.\n2. Use reference images for style/identity consistency only.\n3. Repaint the background according to shot type and angle.\n4. If storyboard conflicts with source text, keep narrative logic from source text.\n5. Keep final visual style consistent with provided references.\n"
},
{
"path": "lib/prompts/novel-promotion/single_panel_image.zh.txt",
"content": "你是一位专业的分镜画师。请根据以下分镜数据生成单张高质量的镜头图片。\n\n【绝对禁止 - 图像中不得出现任何文字 - 最高优先级】\n生成的图像中绝对禁止出现任何文字：\n- 禁止出现镜头类型标签（特写、中景、全景等）\n- 禁止出现镜头运动文字（推、拉、摇等）\n- 禁止出现数字或画面编号（1、2、3等）\n- 禁止出现中文或英文文字\n- 禁止出现水印、注释或符号\n- 参考图上的文字标签仅供你识别使用，禁止画入图中\n- 所有输入信息都是给你的指令，不是要画进图像的内容\n纯视觉内容！纯视觉内容！纯视觉内容！\n- 每个图片只能有一张镜头，禁止一张图片多张镜头，禁止拼图，禁止生成多张图，禁止生成一张图片里面三张图\n\n【⚠️ 画面比例 - 必须严格遵守】\n本次生成的画面比例为：{aspect_ratio}\n- 必须严格按照此比例生成画面\n- 禁止输出与指定比例不符的图像\n- 禁止因参考图比例影响输出比例\n- 每张输出的图片里面只能有一张图，禁止一张图片多张图！\n\n【参考图使用规则】\n- 角色参考：用于参考角色外貌、服装、面部特征、体型\n- 场景参考：仅用于参考环境的构图布局风格和氛围，需根据画面重新绘制，不要直接贴在背景上使用\n- 画面的背景必须根据镜头角度和景别重新绘制\n- 特写/细节镜头应使用虚化或局部背景\n- 参考图上方的文字标签标注了角色/场景名称，请与分镜要求对应\n\n【⚠️ 摄影规则 - 关键】\n如果分镜数据中包含 photography_rules，必须严格遵守：\n- **光照方向**：按照 lighting.direction 的描述绘制光源方向\n- **角色位置**：按照 characters 数组中的 screen_position 确定角色在画面中的位置（左/右/中央）\n- **角色姿势**：按照 characters 数组中的 posture 确定角色姿态（站立/坐着等）\n- **景深**：按照 depth_of_field 的描述控制前后景虚实\n- **色调**：按照 color_tone 的描述确定整体色彩氛围\n\n【分镜内容要求】\n- 根据分镜数据设计画面的视觉内容\n- 确保镜头方向一致（不跳轴），角色位置正确\n- 严格按照文字分镜要求绘制镜头\n\n【⚠️ 原文优先原则 - 重要】\n分镜描述可能存在空间/位置错误。务必与原文交叉验证：\n- 当分镜与原文冲突时：按原文的空间关系、角色位置、动作顺序\n- 参考分镜的：镜头类型、构图、摄影角度\n- 参考原文的：剧情逻辑、空间关系、角色互动、动作顺序\n- 智能结合两者生成最准确的视觉效果\n\n【绝对规则 - 严格遵守】\n- 严格按照分镜要求绘制画面\n- 禁止添加、删除或重排任何镜头\n- 镜头必须与输入完全匹配\n\n【分镜数据】\n{storyboard_text_json_input}\n\n【镜头原文】\n{source_text}\n\n【⚠️ 风格要求 - 必须严格遵守】\n画面风格：{style}\n- 必须严格遵循上传的角色和场景参考图的美术风格\n- 角色绘制风格、线条、色彩必须与角色参考图匹配\n- 环境风格、氛围、色调必须与场景参考图匹配\n- 禁止出现与参考图风格不一致的情况！\n"
},
{
"path": "lib/prompts/novel-promotion/storyboard_edit.en.txt",
"content": "You are an expert storyboard image editor.\nEdit a single panel image or a panel set according to user instruction.\n\nRules:\n1. Do not add any text overlay, subtitle, or technical labels.\n2. If user uploaded reference images, use them as primary visual guidance.\n3. Keep identity and scene continuity unless user requests a change.\n\nUser instruction:\n{user_input}\n\nReturn only the edited image result.\n"
},
{
"path": "lib/prompts/novel-promotion/storyboard_edit.zh.txt",
"content": "你是一个修改编辑图片的大师，你需要根据用户的指令来编辑单个镜头或整组分镜图片，编辑时需要遵守以下规则：\n1:不要添加任何多余标识符文字，如字幕，景别信息等\n2:如果用户上传的有图片，那么就按照图片来进行参考修改\n\n用户的编辑指令如下：{user_input}\n\n请根据指令和原图，输出修改后的图片\n\n\n\n\n\n"
},
{
"path": "lib/prompts/novel-promotion/voice_analysis.en.txt",
"content": "You are a dialogue voice-line analyzer.\nExtract spoken lines from text, assign speaker, estimate emotion intensity, and map to storyboard panels.\n\nOutput format (JSON array only):\n[\n {\n \"lineIndex\": 1,\n \"speaker\": \"Speaker name\",\n \"content\": \"Dialogue line\",\n \"emotionStrength\": 0.3,\n \"matchedPanel\": {\n \"storyboardId\": \"storyboard_id\",\n \"panelIndex\": 0\n }\n }\n]\n\nInput text:\n{input}\n\nCharacter library:\n{characters_lib_name}\n\nCharacter introductions:\n{characters_introduction}\n\nStoryboard JSON:\n{storyboard_json}\n\nRules:\n1. Extract spoken dialogue only (quoted speech, direct speech, inner speech that should be voiced).\n2. Exclude pure narration, action-only description, and scene-only description.\n3. emotionStrength must be between 0.1 and 0.5.\n4. Match panel by order + speaker consistency + semantic relevance.\n5. If no reliable panel match exists, set \"matchedPanel\": null.\n6. Use canonical names from character library when possible.\n7. If there is no spoken dialogue that should be voiced, return [].\n8. Return strict JSON only, no markdown.\n9. ⚠️ JSON SAFETY: All quotation marks in dialogue (\"\"''「」 etc.) MUST be converted to corner brackets「」in JSON string values. NEVER use raw ASCII double quotes \" inside string values.\n"
},
{
"path": "lib/prompts/novel-promotion/voice_analysis.zh.txt",
"content": "你是\"台词发言人分析大师\"。\n任务：从文本中提取需要配音的**对话台词**，分析情绪强度，并匹配对应的视频镜头。\n\n输出格式（只返回JSON，禁止markdown标记）：\n[\n {\n \"lineIndex\": 1,\n \"speaker\": \"发言人名称\",\n \"content\": \"台词内容\",\n \"emotionStrength\": 0.5,\n \"matchedPanel\": {\n \"storyboardId\": \"分镜组ID\",\n \"panelIndex\": 0\n }\n }\n]\n\n分析规则：\n\n1. 【台词提取 - 最重要】\n ✅ 只提取以下类型的内容：\n - **带引号的对话**：\"xxx\" 或 \"xxx\" 或 「xxx」\n - **直接引语**：他说：\"xxx\"、她喊道：\"xxx\"\n - **内心独白**：我心想：\"xxx\"\n \n ❌ 严格排除以下内容：\n - 叙述性文字（无引号的描述）\n - 动作描写（描述角色的动作）\n - 场景描述（描述环境、画面）\n - 章节标题\n - 明确设定为无语言、默片、纯画面表达的内容\n\n ⚠️ 判断标准：这句话是否需要有人\"说出来\"？如果只是描述画面动作，不要提取。\n ⚠️ 如果全文没有任何需要配音的台词，直接返回 []。\n\n2. 【情绪强度 emotionStrength】\n 根据台词的情绪激烈程度，输出0.1-0.5之间的数值（⚠️ 注意：最高不超过0.5，保持语音自然平稳）：\n \n | 情绪类型 | 强度范围 | 示例 |\n |---------|---------|------|\n | 平静/陈述 | 0.1-0.15 | \"好的，我知道了\" |\n | 普通对话 | 0.15-0.2 | \"你今天怎么来了？\" |\n | 疑惑/好奇 | 0.2-0.25 | \"这是怎么回事？\" |\n | 惊讶/意外 | 0.25-0.3 | \"什么？！你说真的？\" |\n | 生气/愤怒 | 0.3-0.35 | \"你给我滚出去！\" |\n | 悲伤/哭泣 | 0.25-0.35 | \"为什么要这样对我...\" |\n | 狂喜/激动 | 0.35-0.4 | \"太好了！我们成功了！\" |\n | 咆哮/嘶吼 | 0.4-0.5 | \"我要杀了你！！！\" |\n\n3. 【发言人识别】\n - 对话内容：识别说话者，如\"他说\"、\"她喊道\"\n - 无引导词的引号内容：根据上下文推断发言人\n - 如果无法确定发言人，设为\"旁白\"\n\n4. 【角色匹配】\n - 角色库：{characters_lib_name}\n - 角色介绍：{characters_introduction}\n - 优先使用角色库中完全一致的名称\n - ⭐ 参考角色介绍理解\"我\"和其他称呼对应的角色\n - 如果不存在，使用原文中的称呼\n\n5. 【镜头匹配 - 严格规则】\n ⚠️ 这是关键步骤，必须严格遵守以下规则：\n \n a) **顺序约束**：\n - 台词在原文中的出现顺序必须与分镜顺序大致对应\n - 第N条台词应该匹配在第N个分镜附近，不能跳跃太远\n - 禁止乱序匹配（如第5条台词匹配到第1个分镜）\n \n b) **发言人校验**：\n - 台词的speaker必须与分镜的characters字段中的角色对应\n - 如果分镜画面角色是\"玄离\"，不能匹配\"柳如烟\"的台词\n - 对话场景：谁说话，就匹配包含说话者的分镜\n \n c) **内容匹配**：\n - 优先匹配台词内容完全包含在分镜text_segment中的情况\n - 其次匹配台词内容与text_segment语义相近的情况\n \n d) **匹配策略**：\n 1. 首先根据text_segment精确匹配台词内容\n 2. 验证分镜角色是否包含台词发言人\n 3. 验证顺序是否合理（前后3个分镜范围内）\n 4. 如果无法满足以上条件，matchedPanel设为null\n \n e) **示例**：\n - 原文顺序：柳如烟说\"殿下身份尊贵\" → 玄离说\"胆子挺大\"\n - 分镜顺序：分镜15(柳如烟特写) → 分镜16(玄离特写)\n - 正确匹配：柳如烟台词→分镜15，玄离台词→分镜16\n - 错误匹配：柳如烟台词→分镜16（发言人不匹配）\n\n6. 【多音字处理 - 重要】\n 为确保TTS语音合成发音正确，对于容易被误读的多音字，需要替换为**读音完全相同（包括声调）的单音字**。\n \n 处理原则：\n a) **识别多音字**：找出台词中的多音字（如：还、行、了、乐、朝、重、都等）\n b) **判断正确读音**：根据上下文语义判断该字在此处的正确读音\n c) **选择替换字**：找一个读音完全相同（声母、韵母、声调都一致）的常用单音字替换\n d) **验证替换**：确保替换后的字读音与原意读音完全一致\n \n 替换示例思路：\n - \"还(huán)给我\" → 用\"环\"替换，因为\"环\"只读huán\n - \"银行(háng)\" → 用\"航\"替换，因为\"航\"只读háng \n - \"了(liǎo)解\" → 用\"聊\"替换，因为\"聊\"只读liáo（接近liǎo）\n - \"快乐(lè)\" → 用\"乐\"保持原字，因为TTS通常能正确读常见词\n - \"重(zhòng)量\" → 用\"众\"替换，因为\"众\"只读zhòng\n \n ⚠️ 注意事项：\n - 必须确保替换字的读音与目标读音完全一致，不要用读音相近但不同的字\n - 例如：\"勒\"读lè或lēi，不能用来替换读le的字\n - 常见词组（如\"了解\"、\"快乐\"、\"音乐\"）TTS通常能正确读，可以保持原字\n - 只有当多音字在特定语境下容易被TTS误读时才需要替换\n\n7. 【JSON安全输出】\n ⚠️ 原文中的所有引号（\"\"''「」等）在 JSON 字符串值中必须统一替换为「」，严禁出现未转义的英文双引号 \"\n - ✅ 正确：\"content\":\"「你好啊」\"\n - ❌ 错误：\"content\":\"\"你好啊\"\"\n\n分镜数据如下：\n{storyboard_json}\n\n原文如下：\n{input}\n"
},
{
"path": "lib/prompts/proxy.ts",
"content": "\nexport async function setProxy() {\n if (process.env.PROXY_URL) { // If you are in China, you must use this proxy:\n const { setGlobalDispatcher, ProxyAgent } = await import(\"undici\");\n const proxyAgent = new ProxyAgent(process.env.PROXY_URL);\n setGlobalDispatcher(proxyAgent);\n }\n }"
},
{
"path": "lib/prompts/skills/api-config-template.system.txt",
"content": "你是 API 配置助手，负责把第三方 API 文档映射为可执行模型模板并保存。\n\n目标：\n1) 最少追问，只收集运行必填字段。\n2) 字段齐全后立刻调用工具保存。\n3) 绝不输出臆测字段；不确定就追问。\n\n工具调用规则：\n1) 单模型保存：saveModelTemplate\n2) 多模型一次配置：saveModelTemplates（优先批量）\n3) compatMediaTemplate 必须是 JSON 对象，不能是 JSON 字符串\n4) 保存前先自检字段名，严禁使用 submit/query 这类旧字段\n\n模板结构必须是：\n{\n \"version\": 1,\n \"mediaType\": \"image\" | \"video\",\n \"mode\": \"sync\" | \"async\",\n \"create\": { \"method\": \"...\", \"path\": \"...\", \"contentType\"?: \"...\", \"bodyTemplate\"?: {}, \"multipartFileFields\"?: [\"...\"] },\n \"status\"?: { \"method\": \"...\", \"path\": \"...\" },\n \"content\"?: { \"method\": \"...\", \"path\": \"...\" },\n \"response\": {\n \"taskIdPath\"?: \"$....\",\n \"statusPath\"?: \"$....\",\n \"outputUrlPath\"?: \"$....\",\n \"outputUrlsPath\"?: \"$....\",\n \"errorPath\"?: \"$....\"\n },\n \"polling\"?: {\n \"intervalMs\": 5000,\n \"timeoutMs\": 600000,\n \"doneStates\": [\"completed\"],\n \"failStates\": [\"failed\", \"error\"]\n }\n}\n\n严格校验约束：\n1) version 只能是 1\n2) mediaType 只能 image/video\n3) mode 只能 sync/async\n4) create.path 必填\n5) create.method 为 POST/PUT/PATCH 时 create.bodyTemplate 必填\n6) mode=async 时 status.path 必填，且必须包含 {{task_id}}\n7) mode=async 时 response.taskIdPath、response.statusPath、polling 必填\n8) mode=sync 时 response.outputUrlPath 或 response.outputUrlsPath 至少一个\n9) JSONPath 字段必须以 $. 开头\n10) 只有 contentType=multipart/form-data 时才能使用 multipartFileFields；且 multipartFileFields 中的字段必须存在于 bodyTemplate 中\n\n最小必问项（只问这些）：\n1) modelId（每个模型）\n2) name（每个模型；无则默认 modelId）\n3) type（image/video）\n4) create endpoint（method + path + body 字段映射）\n5) async 场景下 status endpoint（method + path，且 path 必须带 {{task_id}}）\n6) async 场景下必须明确 taskIdPath、statusPath、polling.doneStates、polling.failStates\n7) 若 create.contentType 是 multipart/form-data，必须明确哪些字段是文件字段，并写入 multipartFileFields\n8) sync 场景下必须明确 outputUrlPath 或 outputUrlsPath\n\n禁止隐式默认：\n1) 不得擅自补 taskIdPath/statusPath/outputUrlPath\n2) 不得擅自补 polling.doneStates/failStates\n3) 文档未明确时必须追问，不能猜\n4) 只有 create.contentType 未写明时，可根据请求示例判断为 application/json 或 multipart/form-data\n\nvideo create bodyTemplate 默认建议（除非文档明确不同）：\n{\n \"model\": \"{{model}}\",\n \"prompt\": \"{{prompt}}\",\n \"seconds\": \"{{duration}}\",\n \"size\": \"{{size}}\",\n \"input_reference\": \"{{image}}\"\n}\n\npath 选择规则（非常重要）：\n1) 若文档给了完整 URL，直接完整保留\n2) 若文档给的是相对路径，原样保留，不要擅自去掉 /v1、/v2 或其他前缀\n3) 不允许自作主张改写版本前缀\n\n交互风格：\n1) 不展示思维链路，不输出 \n2) 先给简短中文结论，再给补充问题或执行结果\n3) 字段齐全时直接调用工具，不要重复确认\n4) 工具返回 invalid 时，逐条复述 issues 并只追问缺失字段，然后再次调用工具\n\n当前 providerId={{providerId}}\n"
},
{
"path": "lib/prompts/skills/tutorial.system.txt",
"content": "你是产品教程助手。\n目标是给出准确、可执行、简洁的操作步骤。\n当用户信息不足时先提问，不要猜测。\n禁止编造不存在的页面、按钮、接口或参数。\n"
},
{
"path": "messages/en/actions.json",
"content": "{\n \"storyboard\": \"Storyboard\",\n \"storyboard_candidate\": \"Storyboard Candidate\",\n \"character\": \"Character\",\n \"location\": \"Location\",\n \"video\": \"Video\",\n \"analyze\": \"Analyze\",\n \"analyze_character\": \"Character Analysis\",\n \"analyze_location\": \"Location Analysis\",\n \"clips\": \"Clip Splitting\",\n \"storyboard_text_plan\": \"Storyboard Planning\",\n \"storyboard_text_detail\": \"Storyboard Details\",\n \"tts\": \"Text-to-Speech\",\n \"regenerate\": \"Regenerate\",\n \"voice-generate\": \"Voice Generation\",\n \"voice-design\": \"Voice Design\",\n \"lip-sync\": \"Lip Sync\"\n}"
},
{
"path": "messages/en/apiConfig.json",
"content": "{\n \"title\": \"API Configuration\",\n \"saving\": \"Saving...\",\n \"saved\": \"Saved\",\n \"saveFailed\": \"Save failed\",\n \"connected\": \"Connected\",\n \"notConfigured\": \"Not configured\",\n \"configure\": \"Configure\",\n \"connect\": \"Connect\",\n \"compatibilityLayerOpenAI\": \"OpenAI Compatible Layer\",\n \"compatibilityLayerGemini\": \"Gemini Compatible Layer\",\n \"show\": \"Show\",\n \"hide\": \"Hide\",\n \"capability\": \"Models\",\n \"default\": \"Default\",\n \"delete\": \"Delete\",\n \"add\": \"Add\",\n \"cancel\": \"Cancel\",\n \"close\": \"Close\",\n \"save\": \"Save\",\n \"comingSoon\": \"Coming soon\",\n \"priceInput\": \"Input {amount}\",\n \"priceOutput\": \"Output {amount}\",\n \"priceUnavailable\": \"N/A\",\n \"fillComplete\": \"Please fill in all fields\",\n \"fillPricing\": \"Please fill in pricing information\",\n \"pricingInputLabel\": \"Input price\",\n \"pricingOutputLabel\": \"Output price\",\n \"pricingBasePriceLabel\": \"Base price (optional)\",\n \"pricingEnableCustom\": \"Enable custom pricing (optional)\",\n \"pricingOptionPricesPlaceholder\": \"Option pricing JSON (optional), e.g. {\\\"resolution\\\":{\\\"1024x1024\\\":0.12}}\",\n \"modelIdExists\": \"Model ID already exists\",\n \"flushConfigFailed\": \"Failed to save provider settings. Please save API key/Base URL first.\",\n \"probeLlmProtocolFailed\": \"Model protocol probe failed. Please try again.\",\n \"probeAuthFailed\": \"Model protocol probe authentication failed. Please check API key.\",\n \"probeInconclusive\": \"Model protocol probe is inconclusive (rate limit or provider error). Please retry later.\",\n \"probeRequestFailed\": \"Model protocol probe request failed. Please try again.\",\n \"modelDisplayName\": \"Display Name (for your reference)\",\n \"modelActualId\": \"Actual Model ID (API parameter)\",\n \"noModelsForProvider\": \"No models configured for this provider\",\n \"defaultModels\": \"Default Model Configuration\",\n \"textDefault\": \"Text Model\",\n \"characterDefault\": \"Character Model\",\n \"locationDefault\": \"Location Model\",\n \"storyboardDefault\": \"Storyboard Model\",\n \"editDefault\": \"Edit Model\",\n \"videoDefault\": \"Video Model\",\n \"audioDefault\": \"Voice Model\",\n \"lipsyncDefault\": \"Lip Sync Model\",\n \"selectDefault\": \"Select\",\n \"defaultModelDesc\": {\n \"analysisModel\": \"Handles script analysis, storyboard construction and full-pipeline text reasoning.\",\n \"videoModel\": \"Synthesizes images and instructions into final video clips.\",\n \"characterModel\": \"Generate character portraits and appearance references from script descriptions\",\n \"locationModel\": \"Generate scene environments and spatial references from script descriptions\",\n \"storyboardModel\": \"Generate shot frames and visual references from storyboard scripts\",\n \"editModel\": \"Perform localized edits, style adjustments and refinements on existing images\",\n \"audioModel\": \"Convert text dialogue into natural and fluent speech audio\",\n \"lipSyncModel\": \"Precisely synchronize speech audio with video character lip movements\",\n \"voiceDesignModel\": \"Design custom voice tones and speech style profiles for characters\"\n },\n \"defaultModelSection\": {\n \"coreFoundation\": \"Text Analysis & Video\",\n \"creativePipeline\": \"Global Image Model Config\",\n \"unifiedOverride\": \"Batch Image Model Config\",\n \"unifiedOverrideHint\": \"Set the model responsible for image generation/editing across the entire system\",\n \"unifiedOverridePlaceholder\": \"Apply to all scenes...\",\n \"followUnified\": \"Follow global config\",\n \"extensions\": \"Extensions\",\n \"coreTextTitle\": \"Text Analysis Model\",\n \"coreVideoTitle\": \"Video Generation Model\",\n \"pipelineCharacter\": \"Character Gen\",\n \"pipelineLocation\": \"Scene Gen\",\n \"pipelineStoryboard\": \"Shot Gen\",\n \"pipelineEdit\": \"Image Edit\",\n \"extLipSync\": \"Lip Sync\",\n \"extTTS\": \"Speech Synthesis\",\n \"extVoiceDesign\": \"Voice Design\",\n \"corePlaceholder\": \"Required\",\n \"extPlaceholder\": \"Not enabled\"\n },\n \"imageModelTip\": \"We recommend using Google's Banana series models. Other image models currently have limited generation quality.\",\n \"customProviderTip\": \"This project is currently in beta. Due to varying custom API formats across providers, custom API compatibility is still limited. We recommend using the built-in official APIs. Future updates will expand compatibility with more providers.\",\n \"providerPool\": \"Provider Pool\",\n \"providerPoolDesc\": \"Configure and use a rich selection of models from global providers\",\n \"dragToSort\": \"Drag to sort\",\n \"dragToSortHint\": \"Drag the top-left handle on each card to reorder providers\",\n \"hideProvider\": \"Hide provider\",\n \"hideProviderConfirm\": \"Are you sure you want to hide this provider? It will be moved to the bottom and can be restored at any time.\",\n \"showProvider\": \"Show provider\",\n \"showHiddenProviders\": \"Show hidden providers\",\n \"hideHiddenProviders\": \"Hide hidden providers\",\n \"hiddenProvidersPrefix\": \"Hidden\",\n \"providerIdExists\": \"Provider ID already exists\",\n \"presetProviderCannotDelete\": \"Preset providers cannot be deleted\",\n \"confirmDeleteProvider\": \"Are you sure you want to delete this provider?\",\n \"presetModelCannotDelete\": \"Preset models cannot be deleted\",\n \"confirmDeleteModel\": \"Are you sure you want to delete this model?\",\n \"addGeminiProvider\": \"Add Model Provider\",\n \"baseUrl\": \"Base URL\",\n \"configureBaseUrl\": \"Configure URL\",\n \"addModel\": \"Add Model\",\n \"batchModeHalfPrice\": \"Batch mode (50% price)\",\n \"openaiCompatVideoOnlyHint\": \"Only OpenAI official-format image-to-video models are supported.\",\n \"typeText\": \"Text\",\n \"typeImage\": \"Image\",\n \"typeVideo\": \"Video\",\n \"typeAudio\": \"Audio\",\n \"apiKeyLabel\": \"API Key\",\n \"apiType\": \"API Type\",\n \"apiTypeGeminiCompatible\": \"Gemini Compatible\",\n \"apiTypeOpenAICompatible\": \"OpenAI Compatible\",\n \"apiTypeGeminiHint\": \"Uses Google SDK\",\n \"otherProviders\": \"Other Settings\",\n \"audioCategory\": \"Audio\",\n \"audioAndLipsync\": \"Audio & Lip Sync\",\n \"configureApiKey\": \"Configure API Key\",\n \"enterApiKey\": \"Enter API Key\",\n \"testConnection\": \"Test Connection\",\n \"testing\": \"Testing...\",\n \"testPassed\": \"Connection test passed\",\n \"testFailed\": \"Connection test failed\",\n \"testWarning\": \"We recommend checking your configuration before adding\",\n \"testRetry\": \"Retry\",\n \"addAnyway\": \"Add Anyway\",\n \"testStep\": {\n \"models\": \"Model List\",\n \"textGen\": \"Text Generation\",\n \"imageGen\": \"Image Generation\",\n \"credits\": \"Credits Check\",\n \"audioGen\": \"Audio Generation\",\n \"skipped\": \"Skipped\"\n },\n \"tabs\": {\n \"llm\": \"Text Models\",\n \"image\": \"Image Models\",\n \"video\": \"Video Models\",\n \"audio\": \"Audio Models\",\n \"other\": \"Other\"\n },\n \"sections\": {\n \"llmApiKeys\": \"Text Model API Keys\",\n \"imageApiKeys\": \"Image Model API Keys\",\n \"videoApiKeys\": \"Video Model API Keys\",\n \"audioApiKey\": \"Audio Model API Key\",\n \"lipsyncApiKey\": \"Lip Sync API Key\"\n },\n \"defaultModel\": {\n \"title\": \"Default Model\",\n \"hint\": \"New projects and Asset Hub will use this default configuration. You can also customize models per project in project settings.\",\n \"notSelected\": \"Not selected\",\n \"analysis\": \"Analysis Model\",\n \"image\": \"Image Generation\",\n \"video\": \"Video Generation\",\n \"resolution\": \"Image Resolution\"\n },\n \"workflowConcurrency\": {\n \"analysis\": \"Analysis Concurrency\",\n \"image\": \"Image Concurrency\",\n \"video\": \"Video Concurrency\"\n },\n \"viewTutorial\": \"View Tutorial\",\n \"tutorial\": {\n \"button\": \"Tutorial\",\n \"title\": \"Setup Guide\",\n \"subtitle\": \"Follow these steps to complete the configuration\",\n \"close\": \"Got it\",\n \"openLink\": \"Open link\",\n \"steps\": {\n \"ark_step1\": \"Go to the Volcano Engine console to create an API Key\",\n \"ark_step2\": \"On the model management page, click 'Enable All Models' button in the top right corner\",\n \"openrouter_step1\": \"Go to OpenRouter platform and create an API Key (must select models with image capabilities)\",\n \"fal_step1\": \"Go to FAL platform and create an API Key\",\n \"google_step1\": \"Go to Google AI Studio and create an API Key\",\n \"minimax_step1\": \"Go to MiniMax platform and get an API Key\",\n \"vidu_step1\": \"Go to the Vidu platform and click 'Create API Key'\",\n \"openai_compatible_step1\": \"Enter any OpenAI-compatible service Base URL and API key\",\n \"gemini_compatible_step1\": \"Enter any Gemini-compatible service Base URL and API key\",\n \"bailian_step1\": \"Go to Alibaba Cloud Bailian console and get an API Key\",\n \"siliconflow_step1\": \"Go to SiliconFlow console and create an API Key\"\n }\n },\n \"assistantOpen\": \"AI Assistant\",\n \"assistantTitle\": \"AI Config Assistant\",\n \"assistantSubtitle\": \"Convert third-party docs into executable image/video templates and auto-save the model.\",\n \"assistantWelcome\": \"Describe your API docs (endpoint, request body, response fields). I will ask follow-up questions and auto-save once valid.\",\n \"assistantInputPlaceholder\": \"Paste docs or describe endpoint details...\",\n \"assistantSend\": \"Send\",\n \"assistantDisabledHint\": \"Save API key and Base URL first\",\n \"assistantRequestFailed\": \"Assistant request failed. Please try again later.\",\n \"assistantResponseInvalid\": \"Assistant response format is invalid. Please retry.\",\n \"assistantMissingTitle\": \"Missing fields\",\n \"assistantWarningsTitle\": \"Warnings\",\n \"assistantDraftTitle\": \"Current draft model\",\n \"assistantReasoningTitle\": \"Reasoning\",\n \"assistantReasoningExpand\": \"Show\",\n \"assistantReasoningCollapse\": \"Hide\",\n \"assistantCompletedTitle\": \"Template Saved\",\n \"assistantCompletedMessage\": \"Model {model} has been added to this provider.\\nClick close to finish this chat.\",\n \"you\": \"You\",\n \"thinking\": \"Thinking...\"\n}\n"
},
{
"path": "messages/en/apiTypes.json",
"content": "{\n \"image\": \"Image Generation\",\n \"video\": \"Video Generation\",\n \"text\": \"Text Analysis\",\n \"tts\": \"Text-to-Speech\",\n \"voice\": \"Voice Dubbing\",\n \"voice_design\": \"Voice Design\",\n \"lip_sync\": \"Lip Sync\"\n}"
},
{
"path": "messages/en/assetHub.json",
"content": "{\n \"title\": \"Asset Hub\",\n \"description\": \"Manage your global character and location assets\",\n \"modelHint\": \"Asset Hub uses default models. To change settings, go to\",\n \"modelHintLink\": \"API Settings\",\n \"modelHintSuffix\": \"\",\n \"folders\": \"Folders\",\n \"noFolders\": \"No folders yet\",\n \"allAssets\": \"All Assets\",\n \"characters\": \"Characters\",\n \"locations\": \"Locations\",\n \"voices\": \"Voices\",\n \"addCharacter\": \"Add Character\",\n \"addLocation\": \"Add Location\",\n \"addVoice\": \"Add Voice\",\n \"downloadAll\": \"Download All\",\n \"downloadAllTitle\": \"Download All Image Assets as ZIP\",\n \"downloading\": \"Packing...\",\n \"downloadSuccess\": \"Download Complete\",\n \"downloadFailed\": \"Download Failed\",\n \"downloadEmpty\": \"No image assets to download\",\n \"newFolder\": \"New Folder\",\n \"editFolder\": \"Edit Folder\",\n \"deleteFolder\": \"Delete Folder\",\n \"folderName\": \"Folder Name\",\n \"folderNamePlaceholder\": \"Enter folder name\",\n \"emptyState\": \"No assets yet\",\n \"emptyStateHint\": \"Click the buttons above to add characters or locations\",\n \"generate\": \"Generate\",\n \"generating\": \"Generating...\",\n \"regenerate\": \"Regenerate\",\n \"undo\": \"Undo\",\n \"delete\": \"Delete\",\n \"cancel\": \"Cancel\",\n \"save\": \"Save\",\n \"create\": \"Create\",\n \"confirmDeleteFolder\": \"Delete this folder? Assets inside will be moved to uncategorized.\",\n \"confirmDeleteCharacter\": \"Delete this character? This action cannot be undone.\",\n \"confirmDeleteLocation\": \"Delete this location? This action cannot be undone.\",\n \"confirmDeleteVoice\": \"Delete this voice? This action cannot be undone.\",\n \"voiceName\": \"Voice Name\",\n \"voiceNamePlaceholder\": \"Enter voice name\",\n \"voiceNameRequired\": \"Please enter a voice name\",\n \"voicePickerTitle\": \"Select from Voice Library\",\n \"voicePickerEmpty\": \"No voices yet. Please create a voice first.\",\n \"voicePickerConfirm\": \"Confirm Selection\",\n \"pagination\": {\n \"previous\": \"Previous\",\n \"next\": \"Next\"\n },\n \"common\": {\n \"cancel\": \"Cancel\"\n },\n \"generateFailed\": \"Generation failed\",\n \"selectFailed\": \"Selection failed\",\n \"uploadFailed\": \"Upload failed\",\n \"editFailed\": \"Edit failed\",\n \"saveVoiceFailed\": \"Failed to save voice\",\n \"saveVoiceFailedDetail\": \"Failed to save voice: {error}\",\n \"bindVoiceFailed\": \"Failed to bind voice\",\n \"bindVoiceFailedDetail\": \"Failed to bind voice: {error}\",\n \"voiceDesignSaved\": \"AI-designed voice has been set for {name}\",\n \"appearanceLabel\": \"Appearance {index}\",\n \"voiceSettings\": {\n \"title\": \"Voice\",\n \"noVoice\": \"No voice\",\n \"previewFailed\": \"Preview failed: {error}\",\n \"uploadFailed\": \"Upload audio failed: {error}\",\n \"uploading\": \"Uploading...\",\n \"uploaded\": \"Uploaded\",\n \"uploadAudio\": \"Upload Audio\",\n \"aiDesign\": \"AI Design\",\n \"voiceLibrary\": \"Voice Library\",\n \"pause\": \"Pause\",\n \"preview\": \"Preview Voice\"\n },\n \"modal\": {\n \"newCharacter\": \"New Character\",\n \"confirm\": \"Confirm\",\n \"processing\": \"Processing...\",\n \"newLocation\": \"New Location\",\n \"addCharacter\": \"Create Character\",\n \"addLocation\": \"Create Location\",\n \"adding\": \"Creating...\",\n \"aiDesign\": \"AI Design\",\n \"aiDesignPlaceholder\": \"e.g., A beautiful woman in a red traditional dress with flowing long hair\",\n \"aiDesignLocationPlaceholder\": \"e.g., A classical Chinese garden with rockery and pavilions\",\n \"aiDesignTip\": \"AI will generate a detailed description based on your input. You can edit it after generation.\",\n \"aiDesignLocationTip\": \"AI will generate a detailed scene description based on your input\",\n \"generate\": \"Generate\",\n \"generating\": \"Generating...\",\n \"nameLabel\": \"Character Name\",\n \"namePlaceholder\": \"Enter character name\",\n \"descLabel\": \"Character Description\",\n \"descPlaceholder\": \"Describe the character's appearance, clothing, hairstyle, etc...\",\n \"locationNameLabel\": \"Location Name\",\n \"locationNamePlaceholder\": \"Enter location name\",\n \"locationSummaryLabel\": \"Location Description\",\n \"locationSummaryPlaceholder\": \"Describe the environment, atmosphere, features, etc...\",\n \"referenceUpload\": \"Upload Reference\",\n \"referenceUploadTip\": \"Upload a character image, AI will convert it to a three-view design sheet\",\n \"convertToCharacter\": \"Convert to 3-View\",\n \"converting\": \"Converting...\",\n \"dropOrClick\": \"Drop image or click to upload\",\n \"supportedFormats\": \"JPG, PNG supported\"\n }\n}"
},
{
"path": "messages/en/assetLibrary.json",
"content": "{\n \"title\": \"Asset Library\",\n \"button\": \"Assets\",\n \"characters\": \"Characters\",\n \"locations\": \"Locations\",\n \"noCharacters\": \"No characters\",\n \"noLocations\": \"No locations\",\n \"addCharacter\": \"Add Character\",\n \"addLocation\": \"Add Location\",\n \"generateImage\": \"Generate Image\",\n \"regenerateImage\": \"Regenerate\",\n \"analyzeAssets\": \"Analyze Assets\",\n \"analyzing\": \"Analyzing...\"\n}"
},
{
"path": "messages/en/assetModal.json",
"content": "{\n \"character\": {\n \"title\": \"New Character\",\n \"name\": \"Character Name\",\n \"namePlaceholder\": \"Enter character name\",\n \"modeReference\": \"Reference Image\",\n \"modeDescription\": \"Description\",\n \"isSubAppearance\": \"This is a sub-appearance\",\n \"isSubAppearanceHint\": \"Add a new appearance state for an existing character\",\n \"uploadReference\": \"Upload Reference\",\n \"pasteHint\": \"Ctrl+V to paste\",\n \"dropOrClick\": \"Click to upload or drag image\",\n \"supportedFormats\": \"Supports JPG, PNG formats\",\n \"nameRequired\": \"Please enter character name first to use reference conversion\",\n \"convertToSheet\": \"Convert to standard character sheet\",\n \"useReferenceGeneratePrefix\": \"Generate from reference\",\n \"generateCountSuffix\": \"images\",\n \"selectReferenceGenerateCount\": \"Select reference generation count\",\n \"referenceTip\": \"Upload any character image, AI will generate a standard character sheet\",\n \"description\": \"Character Description\",\n \"modifyDescription\": \"Modify Description\",\n \"descPlaceholder\": \"Enter character appearance description...\",\n \"modifyDescriptionPlaceholder\": \"Describe how to modify the primary appearance, e.g. formal outfit, post-battle injuries, add a cloak...\",\n \"selectMainCharacter\": \"Select Main Character\",\n \"selectCharacterPlaceholder\": \"Please select a character...\",\n \"appearancesCount\": \"{count} appearances\",\n \"changeReason\": \"Appearance Change Reason\",\n \"changeReasonPlaceholder\": \"e.g. injured after battle, changed into formal wear for a banquet...\",\n \"defaultDescription\": \"{name}'s character profile\",\n \"generationMode\": \"Generation Mode\",\n \"directGenerate\": \"Direct Generate\",\n \"extractPrompt\": \"Extract Prompt\",\n \"extractFirst\": \"Extract Description First\",\n \"directGenerateDesc\": \"Directly generate character sheet from reference (img2img)\",\n \"extractPromptDesc\": \"Extract description from image first, then generate (txt2img)\",\n \"maxReferenceImages\": \"Up to 5 reference images\",\n \"selectedCount\": \"Selected {count}/5 images\",\n \"extractDescription\": \"Extract Description\",\n \"extracting\": \"Extracting...\",\n \"extractedDescription\": \"Extracted Description (Editable)\",\n \"reExtract\": \"Re-extract\",\n \"editHint\": \"Edit the description, then click below to generate\",\n \"generateFromDescription\": \"Generate from Description\",\n \"textToImageTip\": \"Text-to-image mode: Generate from extracted description\",\n \"pleaseExtractFirst\": \"Please extract character description first\"\n },\n \"location\": {\n \"title\": \"New Location\",\n \"name\": \"Location Name\",\n \"namePlaceholder\": \"Enter location name\",\n \"description\": \"Location Description\",\n \"descPlaceholder\": \"Enter location description...\"\n },\n \"artStyle\": {\n \"title\": \"Art Style\"\n },\n \"aiDesign\": {\n \"title\": \"AI Design\",\n \"placeholder\": \"Describe the character you want...\",\n \"placeholderLocation\": \"Describe the scene atmosphere...\",\n \"generating\": \"Designing...\",\n \"generate\": \"Generate\",\n \"tip\": \"Enter a simple description, AI will generate detailed settings\"\n },\n \"common\": {\n \"creating\": \"Creating...\",\n \"create\": \"Create\",\n \"cancel\": \"Cancel\",\n \"adding\": \"Adding...\",\n \"add\": \"Add\",\n \"addOnly\": \"Add character only\",\n \"addOnlyToAssetHub\": \"Add only to asset hub\",\n \"addOnlyLocation\": \"Add location only\",\n \"addOnlyToAssetHubLocation\": \"Add location to asset hub only\",\n \"addAndGeneratePrefix\": \"Add and generate\",\n \"generateCountSuffix\": \"images\",\n \"selectGenerateCount\": \"Select generation count\",\n \"optional\": \"(Optional)\"\n },\n \"errors\": {\n \"uploadFailed\": \"Upload failed\",\n \"extractDescriptionFailed\": \"Failed to extract description\",\n \"createFailed\": \"Creation failed\",\n \"aiDesignFailed\": \"AI design failed\",\n \"addSubAppearanceFailed\": \"Failed to add sub-appearance\",\n \"insufficientBalance\": \"Insufficient balance\"\n }\n}\n"
},
{
"path": "messages/en/assetPicker.json",
"content": "{\n \"selectCharacter\": \"Select Character from Asset Hub\",\n \"selectLocation\": \"Select Location from Asset Hub\",\n \"selectVoice\": \"Select Voice from Asset Hub\",\n \"searchPlaceholder\": \"Search by name or folder...\",\n \"noAssets\": \"No assets in Asset Hub\",\n \"createInAssetHub\": \"Please create characters/locations/voices in Asset Hub first\",\n \"noSearchResults\": \"No matching assets found\",\n \"appearances\": \"appearances\",\n \"images\": \"images\",\n \"cancel\": \"Cancel\",\n \"confirmCopy\": \"Confirm Copy\",\n \"copyFromGlobal\": \"Copy from Asset Hub\",\n \"copySuccess\": \"Copy successful\",\n \"copyFailed\": \"Copy failed\",\n \"preview\": \"Preview\",\n \"stop\": \"Stop\"\n}"
},
{
"path": "messages/en/assets.json",
"content": "{\n \"stage\": {\n \"title\": \"Assets Confirmation\",\n \"characters\": \"Characters\",\n \"locations\": \"Locations\",\n \"analyze\": \"Analyze Assets\",\n \"analyzing\": \"Analyzing...\",\n \"generateAll\": \"Generate All\",\n \"noCharacters\": \"No characters\",\n \"noLocations\": \"No locations\",\n \"confirmProfiles\": \"Character Profiles to Confirm\",\n \"confirmHint\": \"Please confirm these profiles before generating descriptions\",\n \"confirmAll\": \"Confirm All ({count})\",\n \"assetsTitle\": \"Asset Analysis\",\n \"characterAssets\": \"Character Assets\",\n \"locationAssets\": \"Location Assets\",\n \"counts\": \"{characterCount} Characters, {appearanceCount} Appearances\",\n \"locationCounts\": \"{count} Locations\",\n \"undoFailed\": \"Undo failed\",\n \"undoFailedError\": \"Undo failed: {error}\",\n \"undoSuccess\": \"Reverted to previous version\",\n \"editFailed\": \"Edit failed\",\n \"editFailedError\": \"Image edit failed: {error}\",\n \"updateSuccess\": \"Description updated successfully\"\n },\n \"character\": {\n \"add\": \"Add Character\",\n \"edit\": \"Edit Character\",\n \"delete\": \"Delete Character\",\n \"deleteConfirm\": \"Delete this character?\",\n \"deleteAppearanceConfirm\": \"Delete this appearance?\",\n \"deleteFailed\": \"Delete failed: {error}\",\n \"deleteWhole\": \"Delete Whole Character\",\n \"deleteOptions\": \"Delete Options\",\n \"name\": \"Character Name\",\n \"description\": \"Appearance Description\",\n \"generateImage\": \"Generate Profile\",\n \"regenerateImage\": \"Regenerate\",\n \"generate\": \"Generate\",\n \"regenerating\": \"Generating...\",\n \"profile\": \"Profile\",\n \"voiceSettings\": \"Voice Settings\",\n \"speaker\": \"Speaker\",\n \"selectSpeaker\": \"Select Speaker\",\n \"noSpeaker\": \"Not Set\",\n \"primary\": \"Primary\",\n \"secondary\": \"Secondary\",\n \"generateFromPrimary\": \"Generate from Primary\",\n \"selectPrimaryFirst\": \"Select primary first\",\n \"editing\": \"Editing...\",\n \"confirming\": \"Confirming...\",\n \"assetCount\": \"{count} Appearances\",\n \"characterCount\": \"{count} Characters\",\n \"updateFailed\": \"Update description failed\",\n \"addFailed\": \"Add character failed\",\n \"copyFromGlobal\": \"Copy from Asset Hub\"\n },\n \"location\": {\n \"add\": \"Add Location\",\n \"edit\": \"Edit Location\",\n \"delete\": \"Delete Location\",\n \"deleteConfirm\": \"Delete this location?\",\n \"deleteFailed\": \"Delete failed: {error}\",\n \"name\": \"Location Name\",\n \"summary\": \"Summary\",\n \"summaryPlaceholder\": \"Usage/associations, e.g.: John's master bedroom\",\n \"description\": \"Location Description\",\n \"generateImage\": \"Generate Image\",\n \"regenerateImage\": \"Regenerate\",\n \"updateFailed\": \"Update description failed\",\n \"addFailed\": \"Add location failed\"\n },\n \"image\": {\n \"upload\": \"Upload Image\",\n \"uploadReplace\": \"Upload Replacement\",\n \"uploadFailed\": \"Upload Failed\",\n \"uploadFailedError\": \"Upload failed: {error}\",\n \"uploadSuccess\": \"Upload Success!\",\n \"edit\": \"Edit Image\",\n \"editPrompt\": \"Edit Prompt\",\n \"undo\": \"Undo to Previous Version\",\n \"undoConfirm\": \"Are you sure you want to undo to the previous version? Current version will be deleted.\",\n \"regenerateGroup\": \"Regenerate Group\",\n \"regenerateStuck\": \"Click to regenerate (if stuck)\",\n \"selectCount\": \"Select generation count\",\n \"generateCountPrefix\": \"Generate\",\n \"generateCountSuffix\": \"images\",\n \"regenCountPrefix\": \"Regenerate\",\n \"regenCountSuffix\": \"\",\n \"regenCountAriaLabel\": \"Select regeneration count\",\n \"generatedProgress\": \"Generated {generated}/{total}\",\n \"generating\": \"Generating\",\n \"regenerating\": \"Regenerating\",\n \"generatingPlaceholder\": \"Waiting\",\n \"selectTip\": \"Once selected and confirmed, you can edit and modify the image\",\n \"selectFirst\": \"Please select an image first\",\n \"useThis\": \"Use this option\",\n \"optionAlt\": \"{name} - Option {number}\",\n \"optionNumber\": \"Option {number}\",\n \"optionSelected\": \"Selected Option {number}\",\n \"confirmOption\": \"Confirm Option {number}\",\n \"deleteOthersHint\": \"(delete others)\",\n \"confirmSuccess\": \"Selection confirmed\",\n \"confirmFailed\": \"Confirm selection failed: {error}\",\n \"selectFailed\": \"Select image failed: {error}\",\n \"cancelSelection\": \"Cancel Selection\",\n \"deleteThis\": \"Delete this appearance\",\n \"undoFailed\": \"Undo failed\",\n \"undoSuccess\": \"✓ Reverted to previous version\",\n \"editFailed\": \"Image edit failed\",\n \"editSuccess\": \"Image edit successful\",\n \"regenerateFailed\": \"Regenerate failed: {error}\"\n },\n \"modal\": {\n \"newCharacter\": \"New Character\",\n \"addSubAppearance\": \"Add Sub-Appearance\",\n \"aiDesign\": \"AI Design\",\n \"aiDesigning\": \"Designing...\",\n \"designInstruction\": \"Please enter design instruction\",\n \"enterNameDesc\": \"Please enter character name and description\",\n \"selectCharacter\": \"Please select a character\",\n \"enterChangeReason\": \"Please enter change reason\",\n \"enterSubDesc\": \"Please enter appearance description\",\n \"insufficientBalance\": \"Insufficient Balance\\n\\n{error}\",\n \"designFailed\": \"AI Design Failed: {error}\",\n \"addFailed\": \"Add Failed: {error}\",\n \"aiDesignPlaceholderNew\": \"e.g. A 20-year-old female mage, blonde hair, blue eyes...\",\n \"aiDesignPlaceholderSub\": \"e.g. Changed into black combat gear...\",\n \"aiTipNew\": \"Describe the character, AI will generate details\",\n \"aiTipSub\": \"Describe the new state, AI will generate sub-appearance description\",\n \"nameLabel\": \"Character Name\",\n \"namePlaceholder\": \"Enter name...\",\n \"descLabel\": \"Appearance Description\",\n \"descPlaceholder\": \"Enter description...\",\n \"selectLabel\": \"Select Character\",\n \"selectPlaceholder\": \"-- Select Character --\",\n \"existingAppearances\": \"Existing:\",\n \"reasonLabel\": \"Change Reason\",\n \"reasonPlaceholder\": \"e.g. After changing clothes, Injured...\",\n \"reasonTip\": \"Briefly describe the difference from primary appearance\",\n \"subDescPlaceholder\": \"Describe only the changes...\",\n \"subDescTip\": \"Only describe changes (clothes, state), face/body inherits from primary\",\n \"adding\": \"Adding...\",\n \"insufficientBalanceDefault\": \"Insufficient balance, please top up to continue\",\n \"addFailedGeneric\": \"Add Failed\",\n \"appearancesCount\": \"Appearances\",\n \"addCharacter\": \"Add Character\",\n \"addLocation\": \"Add Location\",\n \"aiDesignTip\": \"Describe the scene you want, AI will generate name and details\",\n \"designing\": \"AI designing...\",\n \"saveName\": \"Save Name\",\n \"saveOnly\": \"Save Only\",\n \"sceneDescription\": \"Scene Description\",\n \"scenePrompt\": \"Scene Description Prompt\",\n \"appearancePrompt\": \"Appearance Description Prompt\",\n \"smartModify\": \"Smart Modify\",\n \"modifyPlaceholder\": \"e.g.: Change to night, add moonlight, add curtains...\",\n \"modifyPlaceholderCharacter\": \"e.g.: Change hair to blonde, height to 180cm, wear black suit...\",\n \"modifying\": \"Smart modifying...\",\n \"modifyFailed\": \"Modification failed\",\n \"editCharacter\": \"Edit Character\",\n \"editLocation\": \"Edit Location\",\n \"saveAndGenerate\": \"Save and Generate\",\n \"generatingAutoClose\": \"Generating image, will close automatically when done...\",\n \"aiLocationTip\": \"Enter what you want to modify, AI will adjust the scene description\",\n \"aiDesignPlaceholderLocation\": \"e.g. An ancient magical library, towering bookshelves, dim candlelight, mysterious atmosphere...\",\n \"artStyle\": \"Art Style\",\n \"generate\": \"Generate\",\n \"introduction\": \"Character Introduction\",\n \"introductionPlaceholder\": \"e.g.: The protagonist; 'I' refers to her. Others call her 'Snow' or 'Sister Snow'...\",\n \"introductionTip\": \"Describe the character's role in the story, narrative perspective (who 'I' refers to), how others address them\",\n \"saveIntroduction\": \"Save Introduction\"\n },\n \"toolbar\": {\n \"filter\": \"Filter\",\n \"viewAll\": \"View All\",\n \"showGenerated\": \"Generated\",\n \"showPending\": \"Pending\",\n \"assetManagement\": \"Asset Management\",\n \"assetCount\": \"{total} assets ({appearances} character appearances + {locations} locations)\",\n \"globalAnalyze\": \"Global Analysis\",\n \"globalAnalyzing\": \"Performing global asset analysis...\",\n \"globalAnalyzingHint\": \"Please don't refresh. Results will appear automatically when complete\",\n \"globalAnalyzingTip\": \"Analyzing all episodes, extracting characters and locations...\",\n \"globalAnalyzeHint\": \"Analyze all episodes to extract characters and locations\",\n \"globalAnalyzeSuccess\": \"Global analysis complete: {characters} new characters, {locations} new locations\",\n \"globalAnalyzeFailed\": \"Global analysis failed\",\n \"generateAll\": \"Generate All Images\",\n \"generateAllNoop\": \"All assets already have images, nothing to generate\",\n \"generating\": \"Generating ({current}/{total})\",\n \"regenerateAll\": \"Regenerate All\",\n \"regenerateAllConfirm\": \"Regenerate images for all assets? This will overwrite existing images.\",\n \"noAssetsToGenerate\": \"No assets available for generation\",\n \"regenerateAllHint\": \"Regenerate all asset images (overwrite existing)\",\n \"downloadAll\": \"Download all images as ZIP\"\n },\n \"common\": {\n \"actions\": \"Actions\",\n \"add\": \"Add\",\n \"cancel\": \"Cancel\",\n \"confirm\": \"Confirm\",\n \"copy\": \"Copy\",\n \"delete\": \"Delete\",\n \"download\": \"Download\",\n \"edit\": \"Edit\",\n \"generate\": \"Generate\",\n \"generateFailed\": \"Generation Failed\",\n \"loading\": \"Loading...\",\n \"none\": \"None\",\n \"preview\": \"Preview\",\n \"refresh\": \"Refresh\",\n \"regenerate\": \"Regenerate\",\n \"save\": \"Save\",\n \"status\": \"Status\",\n \"submitFailed\": \"Submit Failed\",\n \"upload\": \"Upload\",\n \"unknownError\": \"Unknown error\"\n },\n \"video\": {\n \"panelCard\": {\n \"generating\": \"Generating...\",\n \"editPrompt\": \"Edit Prompt\"\n }\n },\n \"smartImport\": {\n \"preview\": {\n \"saving\": \"Saving...\"\n }\n },\n \"storyboard\": {\n \"group\": {\n \"generating\": \"Generating...\"\n }\n },\n \"errors\": {\n \"saveFailed\": \"Save Failed, please retry\",\n \"failed\": \"failed, please retry\",\n \"insufficientBalance\": \"Insufficient balance\",\n \"aiDesignFailed\": \"AI design failed\",\n \"createFailed\": \"Creation failed\"\n },\n \"assetLibrary\": {\n \"button\": \"Asset Library\",\n \"title\": \"Asset Library\",\n \"copySuccessCharacter\": \"Character appearance copied successfully\",\n \"copySuccessLocation\": \"Location image copied successfully\",\n \"copySuccessVoice\": \"Voice copied successfully\",\n \"copyFailed\": \"Copy failed: {error}\",\n \"downloadEmpty\": \"No image assets to download\",\n \"downloadFailed\": \"Download failed\"\n },\n \"tts\": {\n \"voiceDesignSaved\": \"AI-designed voice has been set for {name}\",\n \"saveVoiceDesignFailed\": \"Failed to save voice design: {error}\",\n \"title\": \"Voice\",\n \"noVoice\": \"No voice\",\n \"previewFailed\": \"Preview failed: {error}\",\n \"uploadFailed\": \"Upload audio failed: {error}\",\n \"uploadQwenHint\": \"Uploaded voices can only be synthesized with IndexTTS, not QwenTTS. QwenTTS requires an AI-designed voice.\",\n \"uploading\": \"Uploading...\",\n \"uploaded\": \"Uploaded\",\n \"uploadAudio\": \"Upload Audio\",\n \"pause\": \"Pause\",\n \"preview\": \"Preview Voice\"\n },\n \"characterProfile\": {\n \"importance\": {\n \"S\": \"S-Level - Main Protagonist\",\n \"A\": \"A-Level - Core Supporting\",\n \"B\": \"B-Level - Important Supporting\",\n \"C\": \"C-Level - Minor Character\",\n \"D\": \"D-Level - Extra\"\n },\n \"costumeLevel\": {\n \"5\": \"Royal/Luxury\",\n \"4\": \"Noble/Elite\",\n \"3\": \"Professional/Quality\",\n \"2\": \"Casual/Normal\",\n \"1\": \"Plain/Uniform\"\n },\n \"importanceLevel\": \"Character Importance Level\",\n \"characterArchetype\": \"Character Archetype\",\n \"archetypePlaceholder\": \"e.g.: Domineering CEO, Schemer\",\n \"personalityTags\": \"Personality Tags\",\n \"addTagPlaceholder\": \"Add tag\",\n \"costumeLevelLabel\": \"Costume Level\",\n \"suggestedColors\": \"Suggested Colors\",\n \"colorPlaceholder\": \"e.g.: Navy blue, Gold\",\n \"primaryMarker\": \"Primary Identifier\",\n \"markerNote\": \"(Recommended for S/A level)\",\n \"markingsPlaceholder\": \"e.g.: Tear-shaped mole, Silver earring\",\n \"visualKeywords\": \"Visual Keywords\",\n \"keywordsPlaceholder\": \"e.g.: Elite aura, Ascetic style\",\n \"editDialogTitle\": \"Edit Character Profile - {name}\",\n \"confirmAndGenerate\": \"Confirm & Generate\",\n \"useExisting\": \"Use Existing\",\n \"editProfile\": \"Edit Profile\",\n \"delete\": \"Delete Character\",\n \"summary\": {\n \"gender\": \"Gender:\",\n \"age\": \"Age:\",\n \"era\": \"Era:\",\n \"class\": \"Class:\",\n \"occupation\": \"Occupation:\",\n \"personality\": \"Personality:\",\n \"costume\": \"Costume:\",\n \"identifier\": \"Identifier:\"\n },\n \"parseFailed\": \"Failed to parse profile data\",\n \"confirmSuccessGenerating\": \"✓ Profile confirmed. Visual description generation started\",\n \"confirmFailed\": \"Confirm failed: {error}\",\n \"noPendingCharacters\": \"No pending characters to confirm\",\n \"batchConfirmPrompt\": \"Generate visual descriptions for {count} characters?\",\n \"batchConfirmSuccess\": \"✓ Visual descriptions generated for {count} characters\",\n \"batchConfirmFailed\": \"Batch confirmation failed: {error}\",\n \"deleteConfirm\": \"Delete this character? This action cannot be undone.\",\n \"deleteSuccess\": \"✓ Character deleted\",\n \"deleteFailed\": \"Delete failed: {error}\"\n },\n \"imageEdit\": {\n \"editCharacterImage\": \"Edit Character Image\",\n \"editLocationImage\": \"Edit Location Image\",\n \"characterLabel\": \"Character: {name}\",\n \"locationLabel\": \"Location: {name}\",\n \"editInstruction\": \"Edit Instruction\",\n \"subtitle\": \"Enter an edit instruction and optionally upload reference images\",\n \"characterPlaceholder\": \"Describe what you want to change, e.g.: Change hair to blonde, add glasses, change to casual clothes...\",\n \"locationPlaceholder\": \"Describe what you want to change, e.g.: Add more trees, change to night scene...\",\n \"storyboardPlaceholder\": \"Describe what you want to change, e.g.: Change background color, adjust character expression...\",\n \"noAssetHint\": \"No assets, click \\\"Add Asset\\\" to select\",\n \"referenceImages\": \"Reference Images\",\n \"referenceImagesHint\": \"(optional, paste supported)\",\n \"startEditing\": \"Start Editing\"\n }\n}"
},
{
"path": "messages/en/auth.json",
"content": "{\n \"welcomeBack\": \"Welcome Back\",\n \"loginTo\": \"Sign in to waoowaoo\",\n \"createAccount\": \"Create Account\",\n \"joinPlatform\": \"Join waoowaoo\",\n \"phoneNumber\": \"Username\",\n \"password\": \"Password\",\n \"confirmPassword\": \"Confirm Password\",\n \"phoneNumberPlaceholder\": \"Enter your username\",\n \"passwordPlaceholder\": \"Enter your password\",\n \"passwordMinPlaceholder\": \"Enter password (at least 6 characters)\",\n \"confirmPasswordPlaceholder\": \"Re-enter your password\",\n \"loginButton\": \"Sign In\",\n \"loginButtonLoading\": \"Signing in...\",\n \"signupButton\": \"Sign Up\",\n \"signupButtonLoading\": \"Signing up...\",\n \"noAccount\": \"Don't have an account?\",\n \"hasAccount\": \"Already have an account?\",\n \"signupNow\": \"Sign Up Now\",\n \"signinNow\": \"Sign In Now\",\n \"backToHome\": \"← Back to Home\",\n \"loginFailed\": \"Login failed, please check your phone number and password\",\n \"loginError\": \"An error occurred during login\",\n \"passwordMismatch\": \"Passwords do not match\",\n \"passwordTooShort\": \"Password must be at least 6 characters\",\n \"signupSuccess\": \"Registration successful! Redirecting to login page...\",\n \"signupFailed\": \"Registration failed\",\n \"signupError\": \"An error occurred during registration\",\n \"rateLimited\": \"Too many requests, please try again later\",\n \"passwordStrength\": {\n \"weak\": \"Strength: Weak — Use a longer password with uppercase, lowercase, numbers and symbols\",\n \"fair\": \"Strength: Fair — Consider adding more complexity\",\n \"good\": \"Strength: Good\",\n \"strong\": \"Strength: Strong\"\n }\n}"
},
{
"path": "messages/en/billing.json",
"content": "{\n \"transactionType\": \"Transaction Type\",\n \"startDate\": \"Start Date\",\n \"endDate\": \"End Date\",\n \"all\": \"All\",\n \"income\": \"Income\",\n \"expense\": \"Expense\",\n \"reset\": \"Reset\",\n \"filter\": \"Filter\",\n \"noRecords\": \"No records\",\n \"accountRecharge\": \"Account Recharge\",\n \"serviceConsumption\": \"Service Consumption\",\n \"balance\": \"Balance\",\n \"allTypes\": \"All Types\"\n}"
},
{
"path": "messages/en/common.json",
"content": "{\n \"appName\": \"waoowaoo\",\n \"betaVersion\": \"Beta v{version}\",\n \"updateNotice\": {\n \"title\": \"New version available\",\n \"subtitle\": \"Current v{current} · Latest v{latest}\",\n \"description\": \"A newer version has been published on GitHub Releases. View the changelog to update safely.\",\n \"releaseName\": \"Release\",\n \"publishedAt\": \"Published\",\n \"updateTag\": \"Update\",\n \"viewRelease\": \"View update\",\n \"remindLater\": \"Remind me later\",\n \"close\": \"Close\",\n \"openDialog\": \"Open update dialog\",\n \"checkUpdate\": \"Check for updates\",\n \"upToDate\": \"Already up to date\"\n },\n \"loading\": \"Loading...\",\n \"save\": \"Save\",\n \"cancel\": \"Cancel\",\n \"confirm\": \"Confirm\",\n \"delete\": \"Delete\",\n \"edit\": \"Edit\",\n \"search\": \"Search\",\n \"clear\": \"Clear\",\n \"close\": \"Close\",\n \"back\": \"Back\",\n \"next\": \"Next\",\n \"previous\": \"Previous\",\n \"submit\": \"Submit\",\n \"reset\": \"Reset\",\n \"generate\": \"Generate\",\n \"regenerate\": \"Regenerate\",\n \"preview\": \"Preview\",\n \"download\": \"Download\",\n \"upload\": \"Upload\",\n \"select\": \"Select\",\n \"add\": \"Add\",\n \"remove\": \"Remove\",\n \"refresh\": \"Refresh\",\n \"expand\": \"Expand\",\n \"collapse\": \"Collapse\",\n \"all\": \"All\",\n \"none\": \"None\",\n \"success\": \"Success\",\n \"error\": \"Error\",\n \"warning\": \"Warning\",\n \"info\": \"Info\",\n \"copy\": \"Copy\",\n \"paste\": \"Paste\",\n \"apply\": \"Apply\",\n \"autoSave\": \"Auto-save\",\n \"saved\": \"Saved\",\n \"episode\": \"Episode\",\n \"project\": \"Project\",\n \"editEpisodeName\": \"Edit Episode Name\",\n \"deleteEpisode\": \"Delete Episode\",\n \"deleteEpisodeConfirm\": \"Confirm Delete\",\n \"newEpisode\": \"New Episode\",\n \"optional\": \"(Optional)\",\n \"rename\": \"Rename\",\n \"dragToReorder\": \"Drag to reorder\",\n \"episodeNamePlaceholder\": \"Enter episode name...\",\n \"cancelSelection\": \"Cancel selection\",\n \"referenceImage\": \"Reference image\",\n \"previewLarge\": \"Preview large\",\n \"viewOriginal\": \"View original\",\n \"schemeN\": \"Scheme {n}\",\n \"insufficientBalance\": \"Insufficient Balance\",\n \"insufficientBalanceDetail\": \"Insufficient account balance, please recharge to continue\",\n \"operationFailed\": \"Operation failed\",\n \"pleaseRetry\": \"Please retry\",\n \"recommended\": \"Recommended\",\n \"language\": {\n \"select\": \"Select language\",\n \"zh\": \"Chinese\",\n \"en\": \"English\",\n \"switchConfirmTitle\": \"Switch language?\",\n \"switchConfirmMessage\": \"Switching to {targetLanguage} will update not only interface text, but also end-to-end prompts, script generation, and workflow outputs. Continue?\",\n \"switchConfirmAction\": \"Switch now\"\n },\n \"taskStatus\": {\n \"intent\": {\n \"generate\": {\n \"running\": {\n \"image\": \"Generating\",\n \"video\": \"Generating\",\n \"audio\": \"Generating\",\n \"text\": \"Generating\"\n }\n },\n \"regenerate\": {\n \"running\": {\n \"image\": \"Regenerating\",\n \"video\": \"Regenerating\",\n \"audio\": \"Regenerating\",\n \"text\": \"Regenerating\"\n }\n },\n \"modify\": {\n \"running\": {\n \"image\": \"Modifying\",\n \"video\": \"Modifying\",\n \"audio\": \"Modifying\",\n \"text\": \"Modifying\"\n }\n },\n \"analyze\": {\n \"running\": {\n \"image\": \"Analyzing\",\n \"video\": \"Analyzing\",\n \"audio\": \"Analyzing\",\n \"text\": \"Analyzing\"\n }\n },\n \"build\": {\n \"running\": {\n \"image\": \"Building\",\n \"video\": \"Building\",\n \"audio\": \"Building\",\n \"text\": \"Building\"\n }\n },\n \"convert\": {\n \"running\": {\n \"image\": \"Converting\",\n \"video\": \"Converting\",\n \"audio\": \"Converting\",\n \"text\": \"Converting\"\n }\n },\n \"process\": {\n \"running\": {\n \"image\": \"Processing\",\n \"video\": \"Processing\",\n \"audio\": \"Processing\",\n \"text\": \"Processing\"\n }\n }\n },\n \"failed\": {\n \"image\": \"Failed\",\n \"video\": \"Failed\",\n \"audio\": \"Failed\",\n \"text\": \"Failed\"\n }\n }\n}"
},
{
"path": "messages/en/configModal.json",
"content": "{\n \"title\": \"Project Config\",\n \"subtitle\": \"Defaults to the global settings. You can customize models for this project only — changes apply to this project only.\",\n \"saved\": \"Saved\",\n \"autoSave\": \"Auto-save\",\n \"visualStyle\": \"Visual Style\",\n \"modelParams\": \"Model Parameters\",\n \"aspectRatio\": \"Aspect Ratio\",\n \"ttsSettings\": \"TTS Settings\",\n \"loadingModels\": \"Loading model list...\",\n \"analysisModel\": \"Analysis Model\",\n \"characterModel\": \"Character Model\",\n \"locationModel\": \"Location Model\",\n \"storyboardModel\": \"Storyboard Model\",\n \"editModel\": \"Edit Model\",\n \"videoModel\": \"Video Model\",\n \"audioModel\": \"Speech Synthesis Model\",\n \"videoResolution\": \"Video Resolution\",\n \"ttsVoice\": \"TTS Voice\",\n \"ttsRate\": \"Speech Rate\",\n \"fetchModelsFailed\": \"Failed to fetch user model list\",\n \"placeholder\": \"Please enter...\",\n \"description\": \"Description\",\n \"hint\": \"Hint\",\n \"pleaseSelect\": \"Please select...\",\n \"selectModel\": \"Select Model\",\n \"paramConfig\": \"Parameters\",\n \"fixed\": \"Fixed\",\n \"noParams\": \"No configurable parameters\",\n \"confirm\": \"Confirm\",\n \"cancel\": \"Cancel\",\n \"delete\": \"Delete\",\n \"boolOn\": \"On\",\n \"boolOff\": \"Off\"\n}\n"
},
{
"path": "messages/en/errors.json",
"content": "{\n \"UNAUTHORIZED\": \"Please log in first\",\n \"FORBIDDEN\": \"Access denied\",\n \"NOT_FOUND\": \"Resource not found\",\n \"INSUFFICIENT_BALANCE\": \"Insufficient API balance. Please top up and retry\",\n \"RATE_LIMIT\": \"Too many requests. Please retry in {retryAfter} seconds\",\n \"MODEL_NOT_OPEN\": \"Model permission is not activated. Go to https://console.volcengine.com/ark/region:ark+cn-beijing/openManagement?LLM=%7B%7D&advancedActiveKey=model and click \\\"Activate all models\\\" in the top-right of Model Management\",\n \"MODEL_NOT_REGISTERED\": \"Model is not registered. Add an available model in configuration first\",\n \"MODEL_NOT_CONFIGURED\": \"No model configured. Please go to Settings and add the required model type before generating.\",\n \"QUOTA_EXCEEDED\": \"Quota exceeded. Please try again later\",\n \"GENERATION_FAILED\": \"Generation failed. Please retry\",\n \"GENERATION_TIMEOUT\": \"Generation timed out. Please retry\",\n \"SENSITIVE_CONTENT\": \"Content may contain sensitive material\",\n \"INVALID_PARAMS\": \"Invalid parameters\",\n \"MISSING_CONFIG\": \"Please complete model configuration first\",\n \"INTERNAL_ERROR\": \"Server error. Please try again later\",\n \"NETWORK_ERROR\": \"Network error. Please check your connection\",\n \"EMPTY_RESPONSE\": \"Model returned an empty response (no meaningful content). Please retry\",\n \"EXTERNAL_ERROR\": \"External service temporarily unavailable. Please retry later\",\n \"TASK_NOT_READY\": \"Task is still processing\",\n \"NO_RESULT\": \"Task has no result\",\n \"CONFLICT\": \"Resource state conflict\"\n}"
},
{
"path": "messages/en/landing.json",
"content": "{\n \"title\": \"waoowaoo\",\n \"subtitle\": \"AI Film & TV Studio\",\n \"enterWorkspace\": \"Enter Workspace\",\n \"loading\": \"Loading...\",\n \"getStarted\": \"Get Started\",\n \"learnMore\": \"Learn More\",\n \"features\": {\n \"title\": \"Unleash Infinite Creativity\",\n \"subtitle\": \"Full-process AI assistance, from script to final cut\",\n \"character\": {\n \"title\": \"Character Workshop\",\n \"description\": \"Create unique anime characters with high consistency\"\n },\n \"storyboard\": {\n \"title\": \"Smart Storyboard\",\n \"description\": \"One-click text to storyboard, precise narrative control\"\n },\n \"world\": {\n \"title\": \"World Building\",\n \"description\": \"Immersive scene generation to build grand story backgrounds\"\n }\n },\n \"footer\": {\n \"copyright\": \"2026 waoowaoo AI. All rights reserved.\"\n }\n}"
},
{
"path": "messages/en/layout.json",
"content": "{\n \"title\": \"AI Anime Production Platform\",\n \"description\": \"Create professional anime content with cutting-edge AI technology\"\n}\n"
},
{
"path": "messages/en/modelSection.json",
"content": "{\n \"llmModels\": \"Text Model List\",\n \"imageModels\": \"Image Model List\",\n \"videoModels\": \"Video Model List\",\n \"price\": \"Price\",\n \"pricePerMillion\": \"Per million tokens\",\n \"pricePerImage\": \"Per image\",\n \"pricePerVideo\": \"Per video\",\n \"name\": \"Name\",\n \"modelId\": \"Model ID\",\n \"modelName\": \"Model Name\",\n \"provider\": \"Provider\",\n \"resolution\": \"Resolution\",\n \"add\": \"Add\",\n \"addModel\": \"Add Model\",\n \"addNewModel\": \"Add New Model\",\n \"selectPreset\": \"Select Preset Model\",\n \"customModel\": \"Custom Model\",\n \"confirmAdd\": \"Confirm\",\n \"cancel\": \"Cancel\",\n \"done\": \"Done\",\n \"fillComplete\": \"Please fill in all fields\",\n \"noModels\": \"No models yet, click the button above to add\",\n \"noApiKey\": \"Configure API Key\",\n \"batchMode\": \"Batch\",\n \"batchModeTooltip\": \"Offline inference, 50% cheaper, completes within 24 hours\"\n}\n"
},
{
"path": "messages/en/nav.json",
"content": "{\n \"workspace\": \"Workspace\",\n \"assetHub\": \"Asset Hub\",\n \"profile\": \"Settings\",\n \"downloadLogs\": \"Download Logs\",\n \"signin\": \"Sign In\",\n \"signup\": \"Sign Up\",\n \"logout\": \"Logout\"\n}\n"
},
{
"path": "messages/en/novel-promotion.json",
"content": "{\n \"stages\": {\n \"story\": \"Story\",\n \"script\": \"Script\",\n \"storyboard\": \"Storyboard\",\n \"video\": \"Video\",\n \"editor\": \"AI Editor\",\n \"editorComingSoon\": \"Coming soon, follow us for updates\"\n },\n \"buttons\": {\n \"assetLibrary\": \"Asset Library\",\n \"settings\": \"Project Config\",\n \"refreshData\": \"Refresh Data\",\n \"enterVideoGeneration\": \"Enter Video Generation →\"\n },\n \"smartImport\": {\n \"title\": \"Start Your Creative Journey\",\n \"subtitle\": \"First, choose your creation method\",\n \"manualCreate\": {\n \"title\": \"Create from Episode 1\",\n \"description\": \"Start from episode 1, suitable for episodic creation or single short videos\",\n \"button\": \"Start Creating\"\n },\n \"smartImport\": {\n \"title\": \"Smart Import Full Book\",\n \"description\": \"Upload a complete novel or script, AI engine automatically recognizes chapter structure and splits into episodes.\",\n \"button\": \"Import Now\",\n \"recommended\": \"Recommended\"\n },\n \"upload\": {\n \"title\": \"Upload Raw Material\",\n \"subtitle\": \"AI engine is ready, automatic episode splitting and formatting\",\n \"maxWords\": \"(Max 30,000 words)\",\n \"textInput\": \"Enter Text Content\",\n \"documentUpload\": \"Upload Full Document\",\n \"placeholder\": \"Paste your novel chapters or script content here...\",\n \"filePlaceholder\": \"File uploaded mode\",\n \"clickUpload\": \"Click to upload document\",\n \"clearTextFirst\": \"Please clear left text first\",\n \"supportedFormats\": \"Supports Word, TXT formats\",\n \"preview\": \"Preview\",\n \"expandPreview\": \"Expand More\",\n \"collapsePreview\": \"Collapse\",\n \"deleteFile\": \"Delete File\",\n \"startAnalysis\": \"Start Smart Analysis\",\n \"back\": \"Back\",\n \"words\": \"words\"\n },\n \"analyzing\": {\n \"title\": \"AI is Analyzing Your Story\",\n \"description\": \"Recognizing chapter structure, smart splitting in progress...\",\n \"autoSave\": \"Will auto-save after analysis complete\"\n },\n \"preview\": {\n \"title\": \"Smart Splitting Complete\",\n \"episodeCount\": \"Automatically split into {count} episodes\",\n \"totalWords\": \"Total {count} words\",\n \"autoSaved\": \"✓ Auto-saved\",\n \"reanalyze\": \"Re-analyze\",\n \"confirm\": \"Confirm Complete\",\n \"saving\": \"Saving...\",\n \"episodeList\": \"Episode List\",\n \"addEpisode\": \"Add Episode\",\n \"averageWords\": \"Average per episode\",\n \"episodeContent\": \"Episode Content\",\n \"episodePlaceholder\": \"Enter episode title...\",\n \"summaryPlaceholder\": \"Enter plot summary...\",\n \"newEpisode\": \"New Episode\",\n \"deleteEpisode\": \"Delete Episode\",\n \"deleteConfirm\": {\n \"title\": \"Confirm Delete\",\n \"message\": \"Are you sure you want to delete \\\"{title}\\\"?\",\n \"cancel\": \"Cancel\",\n \"confirm\": \"Confirm Delete\"\n },\n \"tip\": {\n \"title\": \"Tip\",\n \"content\": \"You can directly edit titles, summaries, and content. After clicking [Confirm Complete], episodes will be officially imported into the project\"\n }\n },\n \"errors\": {\n \"fileTooLarge\": \"File too large, please upload a file smaller than 10MB\",\n \"docNotSupported\": \".doc format not supported, please convert to .docx in Word\",\n \"fileEmpty\": \"File content is empty\",\n \"fileReadError\": \"File read failed, please try again\",\n \"uploadFirst\": \"Please upload or paste content first\",\n \"analyzeFailed\": \"Analysis failed\",\n \"saveFailed\": \"Save failed\"\n },\n \"cancelConfirm\": \"Are you sure you want to cancel? Analyzed episodes will be cleared.\"\n },\n \"storyInput\": {\n \"currentEditing\": \"Currently editing: {name}\",\n \"editingTip\": \"The following workflow is for this episode only. Switch episodes in the top left if needed\",\n \"wordCount\": \"Word count:\",\n \"assetLibraryTip\": {\n \"title\": \"Need custom characters and locations?\",\n \"description\": \"Click the 「Asset Library」 button in the top right to upload asset setting documents or manually add characters/locations. AI will prioritize using settings from the asset library for analysis.\"\n },\n \"videoRatio\": \"Video Ratio\",\n \"videoRatioHint\": \"Pick a ratio that matches your target platform and content format\",\n \"ratioUsage\": {\n \"1_1\": \"1:1: Square frame, good for avatars, covers and generic social posts\",\n \"9_16\": \"9:16: Vertical video, ideal for TikTok, Reels, Shorts and other short‑video feeds\",\n \"16_9\": \"16:9: Horizontal video, ideal for YouTube, Bilibili and desktop playback\",\n \"4_3\": \"4:3: Classic TV ratio, useful for legacy footage or safe cropping\",\n \"3_4\": \"3:4: Slightly vertical, suitable for mixed text + video layouts\",\n \"2_3\": \"2:3: More vertical, good for posters and character key art\",\n \"3_2\": \"3:2: Slightly horizontal, good for landscapes and story scenes\",\n \"4_5\": \"4:5: Vertical poster ratio, common in social feed images\",\n \"5_4\": \"5:4: Horizontal poster ratio, suitable for PC web banners\",\n \"21_9\": \"21:9: Ultra‑wide cinematic frame, ideal for movie‑like shots and panoramas\"\n },\n \"ratioUsageTag\": {\n \"1_1\": \"Square · Avatars/Covers\",\n \"9_16\": \"Vertical · Short video\",\n \"16_9\": \"Horizontal · Long video\",\n \"4_3\": \"Horizontal · Classic TV\",\n \"3_4\": \"Vertical · Text + video\",\n \"2_3\": \"Vertical · Posters/Key art\",\n \"3_2\": \"Horizontal · Landscape/Story\",\n \"4_5\": \"Vertical · Feed image\",\n \"5_4\": \"Horizontal · Banner\",\n \"21_9\": \"Ultra‑wide · Cinema feel\"\n },\n \"visualStyle\": \"Visual Style\",\n \"visualStyleHint\": \"Pick a style that matches your audience — e.g. Realistic for live‑action, Anime for 2D content\",\n \"currentConfigSummary\": \"Current config: {ratio} · {style}. All subsequent generations will use this combo.\",\n \"assetLibraryRatioNote\": \"Asset library ratios are not affected\",\n \"moreConfig\": \"For more configuration options, click the 「 Settings」 button in the top right\",\n \"narration\": {\n \"title\": \"Enable Narration Voiceover\",\n \"description\": \"Generate TTS voice narration to add commentary to your video\"\n },\n \"creating\": \"AI Creating...\",\n \"ready\": \"✓ Configuration complete, ready for next step\",\n \"pleaseInput\": \"Please enter script content first\"\n },\n \"execution\": {\n \"selectEpisode\": \"Please select an episode first\",\n \"fillContentFirst\": \"Please enter content first\",\n \"requestAborted\": \"Request aborted (possibly due to page refresh)\",\n \"analysisFailed\": \"Asset analysis failed\",\n \"prepareFailed\": \"Preparation failed\",\n \"generationFailed\": \"Generation failed\",\n \"batchVideoFailed\": \"Batch video generation failed\",\n \"updateFailed\": \"Update failed\",\n \"saveFailed\": \"Save failed\",\n \"storyToScriptRunning\": \"Story→Script V2 running\",\n \"scriptToStoryboardRunning\": \"Script→Storyboard V2 running\",\n \"storyToScriptFailed\": \"Story to script failed\",\n \"scriptToStoryboardFailed\": \"Script to storyboard failed\",\n \"taskStreamTimeout\": \"Task timed out. The task may still be running in the background — please check its status or retry\"\n },\n \"rebuildConfirm\": {\n \"storyToScript\": {\n \"title\": \"Script Flow Will Be Rebuilt\",\n \"message\": \"Downstream storyboard data is detected for this episode ({storyboardCount} storyboards, {panelCount} panels). Continuing will clear and rebuild this data. Continue?\"\n },\n \"scriptToStoryboard\": {\n \"title\": \"Storyboard Data Will Be Rebuilt\",\n \"message\": \"Existing storyboard data is detected for this episode ({storyboardCount} storyboards, {panelCount} panels). Continuing will clear current storyboards and regenerate them. Continue?\"\n },\n \"confirm\": \"Continue and Clear\",\n \"cancel\": \"Cancel\"\n }\n}"
},
{
"path": "messages/en/profile.json",
"content": "{\n \"user\": \"User\",\n \"personalAccount\": \"Personal Account\",\n \"availableBalance\": \"Available Balance\",\n \"openSourceNoBilling\": \"Open-source edition, no billing required\",\n \"frozen\": \"Frozen\",\n \"totalSpent\": \"Total Spent\",\n \"apiConfig\": \"API Configuration\",\n \"rechargeRecords\": \"Recharge Records\",\n \"billingRecords\": \"Billing Records\",\n \"logout\": \"Logout\",\n \"downloadLogs\": \"Download Logs\",\n \"accountTransactions\": \"Account Transactions\",\n \"projectDetails\": \"Project Details\",\n \"summary\": \"Summary\",\n \"transactions\": \"Transactions\",\n \"noTransactions\": \"No transaction records\",\n \"noProjectCosts\": \"No project cost records\",\n \"noDetails\": \"This project has no cost details\",\n \"noRecords\": \"No records\",\n \"byType\": \"By Type\",\n \"byAction\": \"By Action\",\n \"times\": \"times\",\n \"total\": \"Total\",\n \"filter\": \"Filter\",\n \"allTypes\": \"All Types\",\n \"recharge\": \"Account Recharge\",\n \"consume\": \"Service Consumption\",\n \"balanceAfter\": \"Balance {amount}\",\n \"recordCount\": \"{count} records\",\n \"totalCost\": \"Total {amount}\",\n \"previousPage\": \"Previous\",\n \"nextPage\": \"Next\",\n \"pagination\": \"{total} items, Page {page} / {totalPages}\",\n \"episodeLabel\": \"Episode {number}\",\n \"billingDetail\": {\n \"imageWithRes\": \"{count} images · {resolution}\",\n \"image\": \"{count} images\",\n \"videoWithRes\": \"{count} videos · {resolution}\",\n \"video\": \"{count} videos\",\n \"tokens\": \"{count} tokens\",\n \"seconds\": \"{count}s\",\n \"calls\": \"{count} calls\"\n },\n \"apiTypes\": {\n \"image\": \"Image Generation\",\n \"video\": \"Video Generation\",\n \"text\": \"Text Analysis\",\n \"tts\": \"Text-to-Speech\",\n \"voice\": \"Voice Acting\",\n \"voice_design\": \"Voice Design\",\n \"lip_sync\": \"Lip Sync\"\n },\n \"actionTypes\": {\n \"image_panel\": \"Storyboard Image\",\n \"image_character\": \"Character Image\",\n \"image_location\": \"Location Image\",\n \"video_panel\": \"Video Generation\",\n \"lip_sync\": \"Lip Sync\",\n \"voice_line\": \"Voice Synthesis\",\n \"voice_design\": \"Voice Design\",\n \"asset_hub_voice_design\": \"Asset Hub Voice Design\",\n \"regenerate_storyboard_text\": \"Regenerate Storyboard Text\",\n \"insert_panel\": \"Insert Panel\",\n \"panel_variant\": \"Shot Variant\",\n \"modify_asset_image\": \"Modify Image\",\n \"regenerate_group\": \"Batch Regenerate\",\n \"asset_hub_image\": \"Asset Hub Image\",\n \"asset_hub_modify\": \"Asset Hub Modify Image\",\n \"analyze_novel\": \"Novel Analysis\",\n \"story_to_script_run\": \"Story to Script\",\n \"script_to_storyboard_run\": \"Script to Storyboard\",\n \"clips_build\": \"Clips Build\",\n \"screenplay_convert\": \"Screenplay Convert\",\n \"voice_analyze\": \"Voice Analysis\",\n \"analyze_global\": \"Global Analysis\",\n \"ai_modify_appearance\": \"AI Modify Appearance\",\n \"ai_modify_location\": \"AI Modify Location\",\n \"ai_modify_shot_prompt\": \"AI Modify Shot Prompt\",\n \"analyze_shot_variants\": \"Analyze Shot Variants\",\n \"ai_create_character\": \"AI Create Character\",\n \"ai_create_location\": \"AI Create Location\",\n \"reference_to_character\": \"Reference to Character\",\n \"character_profile_confirm\": \"Confirm Character Profile\",\n \"character_profile_batch_confirm\": \"Batch Confirm Character Profiles\",\n \"episode_split_llm\": \"Episode Split\",\n \"asset_hub_ai_design_character\": \"Asset Hub AI Design Character\",\n \"asset_hub_ai_design_location\": \"Asset Hub AI Design Location\",\n \"asset_hub_ai_modify_character\": \"Asset Hub AI Modify Character\",\n \"asset_hub_ai_modify_location\": \"Asset Hub AI Modify Location\",\n \"asset_hub_reference_to_character\": \"Asset Hub Reference to Character\",\n \"storyboard\": \"Storyboard\",\n \"storyboard_candidate\": \"Storyboard Candidate\",\n \"character\": \"Character Image\",\n \"location\": \"Location Image\",\n \"video\": \"Video\",\n \"analyze\": \"Analysis\",\n \"analyze_character\": \"Character Analysis\",\n \"analyze_location\": \"Location Analysis\",\n \"clips\": \"Clip Splitting\",\n \"storyboard_text_plan\": \"Storyboard Planning\",\n \"storyboard_text_detail\": \"Storyboard Detail\",\n \"tts\": \"TTS\",\n \"regenerate\": \"Regenerate\",\n \"voice-generate\": \"Voice Generation\",\n \"voice-design\": \"Voice Design\",\n \"lip-sync\": \"Lip Sync\"\n }\n}"
},
{
"path": "messages/en/progress.json",
"content": "{\n \"analyzing\": \"Analyzing story structure...\",\n \"splittingClips\": \"Splitting into clips...\",\n \"convertingScreenplay\": \"Converting to screenplay...\",\n \"submittingStoryboard\": \"Submitting storyboard...\",\n \"step\": \"Step {current} of {total}\",\n \"status\": {\n \"completed\": \"Completed\",\n \"failed\": \"Failed\",\n \"processing\": \"Processing\",\n \"queued\": \"Queued\",\n \"pending\": \"Pending\"\n },\n \"stageCard\": {\n \"stage\": \"Stage\",\n \"realtimeStream\": \"Realtime Stream\",\n \"currentStage\": \"Current Stage\",\n \"outputTitle\": \"Live AI Output · {stage}\",\n \"waitingModelOutput\": \"Waiting for model output...\",\n \"reasoningNotProvided\": \"No reasoning was returned for this step\"\n },\n \"runtime\": {\n \"waitingExecution\": \"Waiting to start\",\n \"taskCreated\": \"Task created\",\n \"taskStarted\": \"Task started\",\n \"taskCompleted\": \"Task completed\",\n \"taskFailed\": \"Task failed\",\n \"taskProcessing\": \"Task processing...\",\n \"llm\": {\n \"processing\": \"Model is processing...\",\n \"output\": \"Model is generating output...\",\n \"reasoning\": \"Model is reasoning...\",\n \"completed\": \"Model output completed\",\n \"failed\": \"Model output failed\"\n },\n \"stage\": {\n \"llmSubmit\": \"Submitting model request\",\n \"llmStreaming\": \"Model streaming output\",\n \"llmFallbackNonStream\": \"Model fallback to non-stream mode\",\n \"llmCompleted\": \"Model output completed\",\n \"llmFailed\": \"Model output failed\"\n }\n },\n \"taskType\": {\n \"generic\": \"Task\",\n \"imagePanel\": \"Storyboard image\",\n \"imageCharacter\": \"Character image\",\n \"imageLocation\": \"Location image\",\n \"videoPanel\": \"Video generation\",\n \"lipSync\": \"Lip sync\",\n \"voiceLine\": \"Voice generation\",\n \"voiceDesign\": \"Voice design\",\n \"assetHubVoiceDesign\": \"Asset hub voice design\",\n \"regenerateStoryboardText\": \"Regenerate storyboard text\",\n \"insertPanel\": \"Insert storyboard panel\",\n \"panelVariant\": \"Storyboard variant\",\n \"modifyAssetImage\": \"Image edit\",\n \"regenerateGroup\": \"Batch regenerate\",\n \"assetHubImage\": \"Asset hub image\",\n \"assetHubModify\": \"Asset hub edit\",\n \"analyzeNovel\": \"Content analysis\",\n \"storyToScriptRun\": \"Story to script\",\n \"scriptToStoryboardRun\": \"Script to storyboard\",\n \"clipsBuild\": \"Clip generation\",\n \"screenplayConvert\": \"Screenplay conversion\",\n \"voiceAnalyze\": \"Voice line analysis\",\n \"analyzeGlobal\": \"Global analysis\",\n \"aiModifyAppearance\": \"Character description modify\",\n \"aiModifyLocation\": \"Location description modify\",\n \"aiModifyShotPrompt\": \"Shot prompt modify\",\n \"analyzeShotVariants\": \"Shot variant analysis\",\n \"aiCreateCharacter\": \"Project character design\",\n \"aiCreateLocation\": \"Project location design\",\n \"referenceToCharacter\": \"Reference to character\",\n \"characterProfileConfirm\": \"Character profile confirm\",\n \"characterProfileBatchConfirm\": \"Character profile batch confirm\",\n \"episodeSplitLlm\": \"Smart episode split\",\n \"assetHubAiDesignCharacter\": \"Asset hub character design\",\n \"assetHubAiDesignLocation\": \"Asset hub location design\",\n \"assetHubAiModifyCharacter\": \"Asset hub character modify\",\n \"assetHubAiModifyLocation\": \"Asset hub location modify\",\n \"assetHubReferenceToCharacter\": \"Asset hub reference to character\"\n },\n \"stage\": {\n \"received\": \"Task received\",\n \"generateCharacterImage\": \"Generate character image\",\n \"generateLocationImage\": \"Generate location image\",\n \"generatePanelCandidate\": \"Generate panel candidate\",\n \"generatePanelVideo\": \"Generate panel video\",\n \"generateVoiceSubmit\": \"Submit voice task\",\n \"generateVoicePersist\": \"Persist voice result\",\n \"voiceDesignSubmit\": \"Submit voice design task\",\n \"voiceDesignDone\": \"Voice design completed\",\n \"submitLipSync\": \"Submit lip sync task\",\n \"persistLipSync\": \"Persist lip sync result\",\n \"storyboardClip\": \"Generate storyboard clip\",\n \"regenerateStoryboardPrepare\": \"Prepare storyboard regeneration\",\n \"regenerateStoryboardPersist\": \"Persist storyboard regeneration\",\n \"storyToScriptPrepare\": \"Prepare story-to-script parameters\",\n \"storyToScriptStep\": \"Execute story-to-script step\",\n \"storyToScriptPersist\": \"Persist story-to-script output\",\n \"storyToScriptPersistDone\": \"Story-to-script output persisted\",\n \"scriptToStoryboardPrepare\": \"Prepare script-to-storyboard parameters\",\n \"scriptToStoryboardStep\": \"Execute script-to-storyboard step\",\n \"scriptToStoryboardPersist\": \"Persist script-to-storyboard output\",\n \"scriptToStoryboardPersistDone\": \"Storyboard and voice output persisted\",\n \"insertPanelGenerateText\": \"Generate inserted panel text\",\n \"insertPanelPersist\": \"Persist inserted panel\",\n \"pollingExternal\": \"Waiting for external service\",\n \"enqueueFailed\": \"Task enqueue failed\",\n \"llmProxySubmit\": \"Submit LLM task\",\n \"llmProxyExecute\": \"Execute LLM task\",\n \"llmProxyPersist\": \"Persist LLM result\"\n },\n \"runConsole\": {\n \"storyToScript\": \"Story to Script\",\n \"scriptToStoryboard\": \"Script to Storyboard\",\n \"storyToScriptRunning\": \"Story→Script running\",\n \"scriptToStoryboardRunning\": \"Script→Storyboard running\",\n \"storyToScriptSubtitle\": \"Story To Script V2\",\n \"scriptToStoryboardSubtitle\": \"Script To Storyboard V2\",\n \"stop\": \"Stop\",\n \"minimize\": \"Minimize\"\n },\n \"streamStep\": {\n \"analyzeCharacters\": \"Analyze characters\",\n \"analyzeLocations\": \"Analyze locations\",\n \"splitClips\": \"Split clips\",\n \"screenplayConversion\": \"Convert screenplay\",\n \"storyboardPlan\": \"Plan storyboard\",\n \"cinematographyRules\": \"Generate cinematography rules\",\n \"actingDirection\": \"Generate acting direction\",\n \"storyboardDetailRefine\": \"Refine storyboard details\",\n \"voiceAnalyze\": \"Analyze voice lines\"\n }\n}\n"
},
{
"path": "messages/en/providerSection.json",
"content": "{\n \"addProvider\": \"+ Add Provider\",\n \"name\": \"Name\",\n \"add\": \"Add\",\n \"save\": \"Save\",\n \"fillRequired\": \"Please fill in required fields\"\n}"
},
{
"path": "messages/en/scriptView.json",
"content": "{\n \"title\": \"Script View\",\n \"scriptBreakdown\": \"Script Breakdown\",\n \"splitCount\": \"{count} clips split\",\n \"noClips\": \"No clips yet, please generate from story view\",\n \"segment\": {\n \"title\": \"Clip {index}\",\n \"selected\": \"(Selected)\"\n },\n \"inSceneAssets\": \"In-Scene Assets\",\n \"currentSelected\": \"Selected: Clip {number}\",\n \"assetView\": {\n \"allClips\": \"All Clips\",\n \"viewingClip\": \"Viewing Clip {number}\"\n },\n \"asset\": {\n \"generateCharacter\": \"Click to generate character →\",\n \"generateLocation\": \"Click to generate location →\",\n \"removeCharacterConfirm\": \"Are you sure you want to remove this character from current clip?\",\n \"removeLocationConfirm\": \"Are you sure you want to remove this location from current clip?\",\n \"removeFromClip\": \"Remove from current clip\",\n \"noAudio\": \"No audio\",\n \"playing\": \"Playing\",\n \"listen\": \"Listen\",\n \"activeCharacters\": \"Active Characters\",\n \"activeLocations\": \"Active Locations\",\n \"selectCharacter\": \"Select character/appearance to add\",\n \"selectLocation\": \"Select location to add\",\n \"loadingAssets\": \"Loading assets...\",\n \"appearanceCount\": \"{count} appearances\",\n \"added\": \"Added\",\n \"primary\": \"Primary\",\n \"subAppearance\": \"Sub appearance\",\n \"defaultAppearance\": \"Default appearance\",\n \"clickToRemove\": \"Click to remove {name}\",\n \"clickToAdd\": \"Click to add {name}\"\n },\n \"screenplay\": {\n \"scene\": \"Scene {number}\",\n \"location\": \"Location:\",\n \"locationTime\": \"Time:\",\n \"day\": \"Day\",\n \"night\": \"Night\",\n \"dawn\": \"Dawn\",\n \"dusk\": \"Dusk\",\n \"dialogue\": \"Dialogue\",\n \"action\": \"Action\",\n \"narration\": \"Narration\",\n \"content\": \"Original Content\",\n \"noContent\": \"No content yet\",\n \"clickToEdit\": \"Click to edit\",\n \"interior\": \"INT\",\n \"exterior\": \"EXT\",\n \"characters\": \"Characters\",\n \"noCharacter\": \"No character info\",\n \"noLocation\": \"No active locations\",\n \"noCharacterInClip\": \"No active characters\"\n },\n \"confirm\": {\n \"removeCharacter\": \"Are you sure you want to remove this character from current clip?\",\n \"removeLocation\": \"Are you sure you want to remove this location from current clip?\"\n },\n \"generate\": {\n \"missingAssets\": \"{count} assets missing images\",\n \"missingAssetsTip\": \"Please generate images for all characters and locations in\",\n \"missingAssetsTipLink\": \"first\",\n \"generating\": \"Generating...\",\n \"startGenerate\": \"Confirm and Start Drawing →\"\n }\n}"
},
{
"path": "messages/en/smartImport.json",
"content": "{\n \"title\": \"Start Your Creative Journey\",\n \"subtitle\": \"First, choose your creation method\",\n \"manualCreate\": {\n \"title\": \"Create from Episode 1\",\n \"description\": \"Start from episode 1, suitable for episodic creation or single short videos\",\n \"button\": \"Start Creating\"\n },\n \"manualDesc\": \"Start from the first episode, suitable for serialized or single short video production\",\n \"startCreate\": \"Start Creating\",\n \"smartImport\": {\n \"title\": \"Smart Text Split\",\n \"description\": \"Upload a complete novel or script, AI engine automatically recognizes chapter structure and splits into episodes.\",\n \"button\": \"Import Now\",\n \"recommended\": \"Recommended\"\n },\n \"markerDetection\": {\n \"enable\": \"Use Markers (Episode X / 第X集)\",\n \"tooltip\": \"Auto-detect [Episode X], [Chapter X], [第X集/章] markers, free & fast\"\n },\n \"smartImportDesc\": \"Upload your novel or script, AI engine automatically identifies chapter structure for one-click smart episode splitting.\",\n \"recommended\": \"Recommended\",\n \"importNow\": \"Import Now\",\n \"uploadTitle\": \"Upload Source Material\",\n \"uploadSubtitle\": \"AI engine ready, one-click auto-split and format\",\n \"maxWords\": \"Max 30,000 words\",\n \"textInput\": \"Enter Text Content\",\n \"textPlaceholder\": \"Paste your novel chapter or script content here...\",\n \"uploadDoc\": \"Upload Complete Document\",\n \"clickUpload\": \"Click to Upload\",\n \"clearText\": \"Please clear left text first\",\n \"supportFormat\": \"Supports Word, TXT formats\",\n \"fileMax\": \"Max 30,000 words\",\n \"words\": \"words\",\n \"startAnalyzing\": \"Start Analysis\",\n \"analyzing\": {\n \"title\": \"AI is Analyzing Your Story\",\n \"description\": \"Recognizing chapter structure, smart splitting in progress...\",\n \"autoSave\": \"Will auto-save after analysis complete\"\n },\n \"analyzingDesc\": \"Identifying chapter structure, smart splitting...\",\n \"autoSave\": \"Will auto-save after analysis\",\n \"splitComplete\": \"Smart Split Complete\",\n \"splitResult\": \"Auto-split into {count} episodes, total {words} words\",\n \"saved\": \"Auto-saved\",\n \"reAnalyze\": \"Re-analyze\",\n \"confirmComplete\": \"Confirm Complete\",\n \"saving\": \"Saving...\",\n \"episodeList\": \"Episode List\",\n \"episodes\": \"episodes\",\n \"episode\": \"Episode {num}\",\n \"addEpisode\": \"Add Episode\",\n \"newEpisode\": \"New Episode\",\n \"avgWords\": \"Average per episode\",\n \"episodeContent\": \"Episode Content\",\n \"plotSummary\": \"Plot Summary\",\n \"enterTitle\": \"Enter episode title...\",\n \"enterSummary\": \"Enter plot summary...\",\n \"confirmDelete\": \"Confirm Delete\",\n \"deleteConfirmMsg\": \"Are you sure you want to delete \\\"{title}\\\"?\",\n \"preview\": {\n \"title\": \"Smart Splitting Complete\",\n \"episodeCount\": \"Automatically split into {count} episodes\",\n \"totalWords\": \"Total {count} words\",\n \"autoSaved\": \"✓ Auto-saved\",\n \"reanalyze\": \"Re-analyze\",\n \"confirm\": \"Confirm Complete\",\n \"saving\": \"Saving...\",\n \"episodeList\": \"Episode List\",\n \"addEpisode\": \"Add Episode\",\n \"averageWords\": \"Average per episode\",\n \"episodeContent\": \"Episode Content\",\n \"episodePlaceholder\": \"Enter episode title...\",\n \"summaryPlaceholder\": \"Enter plot summary...\",\n \"newEpisode\": \"New Episode\",\n \"deleteEpisode\": \"Delete Episode\",\n \"deleteConfirm\": {\n \"title\": \"Confirm Delete\",\n \"message\": \"Are you sure you want to delete \\\"{title}\\\"?\",\n \"cancel\": \"Cancel\",\n \"confirm\": \"Confirm Delete\"\n },\n \"tip\": {\n \"title\": \"Tip\",\n \"content\": \"You can directly edit titles, summaries, and content. After clicking [Confirm Complete], episodes will be officially imported into the project\"\n }\n },\n \"collapsePreview\": \"Collapse Preview\",\n \"expandMore\": \"Expand More\",\n \"deleteFile\": \"Delete File\",\n \"fileTooLarge\": \"File size cannot exceed 10MB\",\n \"docNotSupported\": \".doc format not supported. Please save as .docx or .txt and try again\",\n \"fileEmpty\": \"File content is empty\",\n \"fileReadError\": \"File read failed, please ensure correct format\",\n \"uploadFirst\": \"Please upload a file or paste text first\",\n \"analyzeFailed\": \"Analysis failed\",\n \"saveFailed\": \"Save failed\",\n \"cancelConfirm\": \"Are you sure you want to cancel? Analyzed episodes will be cleared.\",\n \"deleteEpisode\": \"Delete Episode\",\n \"upload\": {\n \"title\": \"Upload Raw Material\",\n \"subtitle\": \"AI engine is ready, automatic episode splitting and formatting\",\n \"maxWords\": \"(Max 30,000 words)\",\n \"textInput\": \"Enter Text Content\",\n \"documentUpload\": \"Upload Full Document\",\n \"placeholder\": \"Paste your novel chapters or script content here...\",\n \"filePlaceholder\": \"File uploaded mode\",\n \"clickUpload\": \"Click to upload document\",\n \"clearTextFirst\": \"Please clear left text first\",\n \"supportedFormats\": \"Supports Word, TXT formats\",\n \"preview\": \"Preview\",\n \"expandPreview\": \"Expand More\",\n \"collapsePreview\": \"Collapse\",\n \"deleteFile\": \"Delete File\",\n \"startAnalysis\": \"Start Analysis\",\n \"back\": \"Back\",\n \"words\": \"words\"\n },\n \"errors\": {\n \"fileTooLarge\": \"File too large, please upload a file smaller than 10MB\",\n \"docNotSupported\": \".doc format not supported, please convert to .docx in Word\",\n \"fileEmpty\": \"File content is empty\",\n \"fileReadError\": \"File read failed, please try again\",\n \"uploadFirst\": \"Please upload or paste content first\",\n \"analyzeFailed\": \"Analysis failed\",\n \"saveFailed\": \"Save failed\",\n \"analysisModelNotConfigured\": \"Please configure an analysis model in settings first\"\n },\n \"common\": {\n \"edit\": \"Edit\",\n \"delete\": \"Delete\",\n \"save\": \"Save\",\n \"cancel\": \"Cancel\"\n },\n \"markerDetected\": {\n \"title\": \"Episode Markers Detected\",\n \"description\": \"Detected {count} \\\"{type}\\\" format episode markers\",\n \"preview\": \"Preview Split Result\",\n \"useMarker\": \"Use Marker Split\",\n \"useMarkerDesc\": \"Fast & Free\",\n \"useAI\": \"Use AI Smart Split\",\n \"useAIDesc\": \"Intelligent analysis, uses credits\",\n \"cancel\": \"Cancel\",\n \"totalCount\": \"{count} episodes total\",\n \"markerTypes\": {\n \"episode\": \"Episode X (Chinese)\",\n \"chapter\": \"Chapter X (Chinese)\",\n \"act\": \"Act X (Chinese)\",\n \"scene\": \"X-Y [Scene]\",\n \"numbered\": \"Numbered\",\n \"numberedEscaped\": \"Numbered (Escaped)\",\n \"numberedDirect\": \"Number + Chinese\",\n \"episodeEn\": \"Episode X\",\n \"chapterEn\": \"Chapter X\",\n \"boldNumber\": \"**Number**\",\n \"pureNumber\": \"Pure Number\"\n }\n },\n \"globalAnalysis\": {\n \"title\": \"Global Asset Analysis\",\n \"description\": \"Extract all characters and locations from the full book to ensure consistency across episodes\",\n \"startButton\": \"Analyze Now\",\n \"analyzing\": \"Analyzing...\",\n \"success\": \"Analysis complete: {characters} new characters, {locations} new locations\",\n \"failed\": \"Global analysis failed\",\n \"confirmAndAnalyze\": \"Confirm & Analyze Assets\"\n }\n}"
},
{
"path": "messages/en/stages.json",
"content": "{\n \"config\": \"1. Config\",\n \"assets\": \"2. Asset Analysis\",\n \"storyboard\": \"3. Storyboard Edit\",\n \"videos\": \"4. Video Generation\",\n \"voice\": \"5. Voice Generation\"\n}"
},
{
"path": "messages/en/storyboard.json",
"content": "{\n \"phases\": {\n \"planning\": \"Planning Storyboard\",\n \"cinematography\": \"Cinematography Design\",\n \"acting\": \"Acting Direction\",\n \"detail\": \"Adding Details\"\n },\n \"prompts\": {\n \"imagePrompt\": \"Image Prompt\",\n \"aiInstruction\": \"AI Modify Instruction\",\n \"supportReference\": \"(Support @ referencing asset library)\",\n \"instructionPlaceholder\": \"e.g. Change location to @Hospital_Day, character to @ProtagonistA\",\n \"selectAsset\": \"Select Asset\",\n \"character\": \"Character\",\n \"location\": \"Location\",\n \"referencedAssets\": \"Referenced Assets:\",\n \"removeAsset\": \"Remove Asset\",\n \"aiModify\": \"AI Modify & Generate\",\n \"aiModifying\": \"Modifying...\",\n \"aiModifyTip\": \"Click to auto-save prompt and generate new image\",\n \"save\": \"Save\",\n \"currentPrompt\": \"Current Prompt\",\n \"enterInstruction\": \"Please enter instruction\",\n \"modifyFailed\": \"Operation Failed: {error}\",\n \"updateFailed\": \"Update Failed: {error}\",\n \"enterContinuation\": \"Please enter content to append\",\n \"appendTitle\": \"Continue Content\",\n \"appendDescription\": \"Enter new SRT content. The system will split and generate new shots, then append them to the end.\",\n \"appendSubmit\": \"Append and Generate Shots\",\n \"appendSuccess\": \"Append succeeded. New shots were added to the end of the list.\",\n \"appendFailed\": \"Append failed: {error}\",\n \"customStyle\": \"Custom Style\"\n },\n \"group\": {\n \"generating\": \"Generating...\",\n \"hasSynced\": \"✓ Generated\",\n \"failed\": \"Failed\",\n \"retry\": \"Retry\",\n \"regenerate\": \"Regenerate All\",\n \"generateAll\": \"Generate All\",\n \"expand\": \"Expand\",\n \"collapse\": \"Collapse\",\n \"addPanel\": \"Add Panel\",\n \"regenerating\": \"Regenerating...\",\n \"aiAnalyzing\": \"AI Analyzing...\",\n \"regenerateText\": \"Regenerate Text\",\n \"generateMissingImages\": \"Generate all panels without images in this segment\",\n \"segment\": \"Segment\",\n \"addAtStart\": \"Add new storyboard group at the start\",\n \"insertHere\": \"Insert new storyboard group here\"\n },\n \"header\": {\n \"title\": \"Storyboard Editing\",\n \"panels\": \"Panels\",\n \"submit\": \"Submit Generation\",\n \"submitting\": \"Submitting...\",\n \"storyboardPanel\": \"Storyboard Panel\",\n \"segments\": \"segments\",\n \"segmentsCount\": \"Total {count} segments,\",\n \"panelsCount\": \"{count} panels\",\n \"generatingStatus\": \"({count} generating)\",\n \"generateAllPanels\": \"Generate All Panels\",\n \"generatePendingPanels\": \"Generate {count} panels without images\",\n \"downloadAll\": \"Download All\",\n \"downloading\": \"Packing...\",\n \"noImages\": \"No images to download\",\n \"downloadAllImages\": \"Download all images\",\n \"generateVideo\": \"Generate Video →\",\n \"back\": \"← Back\",\n \"concurrencyLimit\": \"Concurrency limit {count}\"\n },\n \"panel\": {\n \"shotType\": \"Shot Type:\",\n \"duration\": \"seconds\",\n \"location\": \"Location:\",\n \"characters\": \"Characters:\",\n \"description\": \"Description:\",\n \"text\": \"Corresponding Text:\",\n \"regenerate\": \"Regenerate\",\n \"delete\": \"Delete\",\n \"insertBefore\": \"Insert Before\",\n \"insertAfter\": \"Insert After\",\n \"moveUp\": \"Move Up\",\n \"moveDown\": \"Move Down\",\n \"plot\": \"Plot:\",\n \"summary\": \"Summary:\",\n \"pov\": \"POV:\",\n \"focus\": \"Focus:\",\n \"mode\": \"Mode:\",\n \"shot\": \"Shot\",\n \"segment\": \"Segment\",\n \"stylePrompt\": \"Style/Prompt\",\n \"shotMode\": \"Shot/Mode\",\n \"regenerateImage\": \"Regenerate Image\",\n \"generateImage\": \"Generate Image\",\n \"cardView\": \"Card View\",\n \"tableView\": \"Table View\",\n \"shotTypeLabel\": \"Shot Type\",\n \"cameraMove\": \"Camera Move\",\n \"sourceText\": \"Source Text\",\n \"sceneDescription\": \"Scene Description\",\n \"videoPrompt\": \"Video Prompt\",\n \"videoPromptHint\": \"Describe subject movement, environment, and camera language\",\n \"locationLabel\": \"Location\",\n \"editLocation\": \"Edit Location\",\n \"characterLabel\": \"Character\",\n \"characterLabelWithCount\": \"Characters ({count})\",\n \"editCharacter\": \"Edit Characters\",\n \"select\": \"+ Select\",\n \"add\": \"+ Add\",\n \"noLocation\": \"No location selected\",\n \"locationNotEdited\": \"Location not edited yet\",\n \"noCharacters\": \"No characters selected\",\n \"charactersNotEdited\": \"Characters not edited yet\",\n \"shotTypePlaceholder\": \"Overhead medium shot...\",\n \"cameraMovePlaceholder\": \"Slow push, static...\",\n \"videoPromptPlaceholder\": \"Prompt for video generation...\",\n \"sceneDescriptionPlaceholder\": \"Describe subject, composition, lighting, and mood\",\n \"selectCharacter\": \"Select Character\",\n \"selectLocation\": \"Select Location\",\n \"noCharacterAssets\": \"No character assets\",\n \"noLocationAssets\": \"No location assets\",\n \"selected\": \"Selected\",\n \"defaultAppearance\": \"Default appearance\",\n \"newPanelDescription\": \"New shot description\",\n \"noShotType\": \"Shot type not set\"\n },\n \"image\": {\n \"generating\": \"Generating...\",\n \"regenerate\": \"Regenerate\",\n \"edit\": \"Edit\",\n \"editImage\": \"Edit Image\",\n \"candidate\": \"Candidate\",\n \"selectCandidate\": \"Select Candidate\",\n \"variants\": \"Variants\",\n \"generateVariants\": \"Generate Variants\",\n \"forceRegenerate\": \"Force Regenerate\",\n \"failed\": \"Generation Failed\",\n \"clickToPreview\": \"Click to preview\",\n \"enlargePreview\": \"Enlarge Preview\",\n \"candidateCount\": \"Candidate {count}\",\n \"candidateGenerating\": \"{count} generating\",\n \"selectingCandidate\": \"Selecting candidate...\",\n \"confirmCandidate\": \"Confirm Selection\",\n \"cancelSelection\": \"Cancel Selection\",\n \"noValidCandidates\": \"No valid candidates\",\n \"selectCount\": \"Select count\",\n \"generateMultiple\": \"Generate multiple candidates\",\n \"generateCount\": \"Generate {count}\",\n \"generateCountSuffix\": \"candidates\",\n \"undoShort\": \"Back\"\n },\n \"candidate\": {\n \"title\": \"Select Candidate Image\",\n \"select\": \"Select\",\n \"cancel\": \"Cancel\",\n \"noImages\": \"No candidate images\",\n \"original\": \"Original\"\n },\n \"variant\": {\n \"title\": \"Image Variants\",\n \"generate\": \"Generate Variants\",\n \"select\": \"Use This Image\",\n \"close\": \"Close\",\n \"shotTitle\": \"Shot Variant - Based on #{number}\",\n \"originalDescription\": \"Original Shot Description\",\n \"noDescription\": \"No description\",\n \"noImage\": \"No image\",\n \"shotNum\": \"Shot {number}\",\n \"aiRecommend\": \"AI Recommended Variants\",\n \"reanalyze\": \"Re-analyze\",\n \"shotType\": \"Shot type:\",\n \"cameraMove\": \"Camera move:\",\n \"generating\": \"Generating\",\n \"clickToAnalyze\": \"Click Re-analyze to get AI recommendations\",\n \"customInstruction\": \"Or custom instruction\",\n \"customPlaceholder\": \"Enter the shot effect you want, e.g.: switch to reverse shot, focus on another character's expression...\",\n \"includeCharacter\": \"Include character reference\",\n \"includeLocation\": \"Include location reference\",\n \"customVariant\": \"Custom variant\",\n \"defaultShotType\": \"Medium Shot\",\n \"defaultCameraMove\": \"Static\",\n \"useCustomGenerate\": \"Generate with custom\",\n \"analyzeFailed\": \"Analysis failed\",\n \"creativeScore\": \"Creativity {score}/5\"\n },\n \"insert\": {\n \"title\": \"Insert New Panel\",\n \"position\": \"Insert Position\",\n \"before\": \"Before Panel {number}\",\n \"after\": \"After Panel {number}\",\n \"content\": \"Panel Content\",\n \"shotType\": \"Shot Type\",\n \"location\": \"Location\",\n \"characters\": \"Characters\",\n \"description\": \"Description\",\n \"text\": \"Corresponding Text\",\n \"placeholder\": {\n \"shotType\": \"Select shot type...\",\n \"location\": \"Enter location...\",\n \"characters\": \"Enter characters, comma separated\",\n \"description\": \"Describe the scene...\",\n \"text\": \"Corresponding script text...\"\n },\n \"insert\": \"Insert\",\n \"cancel\": \"Cancel\"\n },\n \"common\": {\n \"actions\": \"Actions\",\n \"add\": \"Add\",\n \"cancel\": \"Cancel\",\n \"confirm\": \"Confirm\",\n \"copy\": \"Copy\",\n \"delete\": \"Delete\",\n \"download\": \"Download\",\n \"edit\": \"Edit\",\n \"generate\": \"Generate\",\n \"loading\": \"Loading...\",\n \"none\": \"None\",\n \"unknownError\": \"Unknown error\",\n \"preview\": \"Preview\",\n \"refresh\": \"Refresh\",\n \"regenerate\": \"Regenerate\",\n \"deleting\": \"Deleting\",\n \"editing\": \"Editing\",\n \"saving\": \"Saving...\",\n \"saveFailed\": \"Save failed, changes not synced\",\n \"retrySave\": \"Retry save\",\n \"save\": \"Save\",\n \"status\": \"Status\",\n \"submitFailed\": \"Submit Failed\",\n \"upload\": \"Upload\"\n },\n \"confirm\": {\n \"deletePanel\": \"Delete this shot? This action cannot be undone.\",\n \"deleteGroup\": \"Delete this storyboard group? This will remove all {count} shots in this segment. This action cannot be undone.\"\n },\n \"messages\": {\n \"episodeNotFound\": \"Episode information not found\",\n \"downloadFailed\": \"Download failed: {error}\",\n \"panelNotFound\": \"Shot information not found\",\n \"modifyFailed\": \"Modify failed: {error}\",\n \"selectCandidateFailed\": \"Select candidate failed: {error}\",\n \"insertPanelFailed\": \"Insert shot failed: {error}\",\n \"addPanelFailed\": \"Add shot failed: {error}\",\n \"deletePanelFailed\": \"Delete shot failed: {error}\",\n \"deleteGroupFailed\": \"Delete storyboard group failed: {error}\",\n \"regenerateGroupFailed\": \"Regenerate storyboard failed: {error}\",\n \"addGroupFailed\": \"Add storyboard group failed: {error}\",\n \"moveGroupFailed\": \"Move storyboard group failed: {error}\",\n \"batchGenerateCompleted\": \"Batch generation completed:\\nSucceeded: {succeeded}\\nFailed: {failed}\\n\\nSample errors: {errors}\",\n \"batchGenerateFailed\": \"Batch generation failed: {error}\"\n },\n \"canvas\": {\n \"emptyTitle\": \"No storyboard data yet\",\n \"emptyDescription\": \"Generate clips and storyboard text first, or add a storyboard group above\"\n },\n \"imageEdit\": {\n \"title\": \"Edit Storyboard Image\",\n \"subtitle\": \"Enter a modify instruction and optionally upload reference images or assets\",\n \"promptPlaceholder\": \"Describe what to modify, e.g. change background color or adjust expression...\",\n \"referenceImagesLabel\": \"Reference Images\",\n \"referenceImagesHint\": \"(optional, paste supported)\",\n \"start\": \"Start Editing\",\n \"selectAsset\": \"Select Assets\",\n \"selectedAssetsLabel\": \"Referenced Assets\",\n \"selectedAssetsCount\": \"{count}\",\n \"addAsset\": \"Add Asset\",\n \"noAssets\": \"No assets selected. Click \\\"Add Asset\\\" to choose.\"\n },\n \"screenplay\": {\n \"tabs\": {\n \"formatted\": \"Screenplay\",\n \"original\": \"Original\"\n },\n \"scene\": \"Scene {number}\",\n \"characters\": \"Characters\",\n \"voiceover\": \"Voiceover\",\n \"parseFailedTitle\": \"Failed to parse screenplay format\",\n \"parseFailedDescription\": \"Please check the original content\"\n },\n \"assets\": {\n \"character\": {\n \"confirming\": \"Confirming...\",\n \"editing\": \"Editing...\"\n },\n \"image\": {\n \"undo\": \"Undo to Previous Version\"\n },\n \"location\": {\n \"generateImage\": \"Generate Image\"\n },\n \"stage\": {\n \"analyzing\": \"Analyzing...\"\n }\n },\n \"video\": {\n \"toolbar\": {\n \"showPending\": \"Pending\"\n },\n \"panelCard\": {\n \"forceRegenerate\": \"Force Regenerate (if stuck)\"\n }\n },\n \"smartImport\": {\n \"errors\": {\n \"analyzeFailed\": \"Analysis Failed\"\n },\n \"preview\": {\n \"reanalyze\": \"Re-analyze\"\n },\n \"smartImport\": {\n \"recommended\": \"Recommended\"\n }\n },\n \"aiData\": {\n \"title\": \"AI Data Editor\",\n \"subtitle\": \"Panel {number} - Complete data sent to image generation AI\",\n \"basicData\": \"Storyboard Basic Data\",\n \"shotType\": \"Shot Type\",\n \"cameraMove\": \"Camera Movement\",\n \"shotTypePlaceholder\": \"Overhead, wide shot, eye-level, medium shot...\",\n \"cameraMovePlaceholder\": \"Slow push, static, follow...\",\n \"scene\": \"Scene (Read-only)\",\n \"notSelected\": \"Not selected\",\n \"summary\": \"Scene Summary\",\n \"characters\": \"Characters (Read-only)\",\n \"plot\": \"Plot\",\n \"summarize\": \"Summary\",\n \"visualDescription\": \"Visual Description\",\n \"videoPrompt\": \"Video Prompt\",\n \"negativePrompt\": \"Negative Prompt\",\n \"save\": \"Save\",\n \"cancel\": \"Cancel\",\n \"lightingDirection\": \"Lighting Direction\",\n \"lightingQuality\": \"Lighting Quality\",\n \"depthOfField\": \"Depth of Field\",\n \"colorTone\": \"Color Tone\",\n \"characterPosition\": \"Character Position Rules\",\n \"position\": \"Position\",\n \"posture\": \"Posture\",\n \"facing\": \"Facing\",\n \"photographyRules\": \"Photography Rules\",\n \"viewData\": \"View Data\",\n \"jsonPreview\": \"JSON Preview\",\n \"actingNotes\": \"Acting Direction (acting_notes)\",\n \"actingTitle\": \"Acting Direction\",\n \"actingDescription\": \"Performance Notes\",\n \"noActingData\": \"No acting data\"\n },\n \"insertModal\": {\n \"insertBetween\": \"Insert between #{before} and #{after}\",\n \"panel\": \"Panel\",\n \"noImage\": \"No image\",\n \"insertAtEnd\": \"End\",\n \"aiAnalyze\": \"AI Auto-analyze\",\n \"analyzing\": \"AI analyzing...\",\n \"insert\": \"Insert\",\n \"inserting\": \"Inserting...\",\n \"placeholder\": \"Optional: Add notes, e.g. add a reaction shot...\"\n },\n \"panelActions\": {\n \"insertPanel\": \"Insert Panel\",\n \"panelVariant\": \"Panel Variant\",\n \"insertHere\": \"Insert panel here\",\n \"generateVariant\": \"Generate variant based on this panel\",\n \"needImage\": \"Need to generate image first\",\n \"deleteShot\": \"Delete Shot\",\n \"pasteSrtPlaceholder\": \"Paste new SRT content...\"\n },\n \"firstLastFrame\": {\n \"placeholder\": \"Enter first/last frame video prompt...\",\n \"modelTitle\": \"First/Last Frame Model\"\n }\n}\n"
},
{
"path": "messages/en/video.json",
"content": "{\n \"panelCard\": {\n \"play\": \"Play\",\n \"pause\": \"Pause\",\n \"retry\": \"Retry\",\n \"regenerate\": \"Regenerate\",\n \"download\": \"Download\",\n \"edit\": \"Edit\",\n \"save\": \"Save\",\n \"cancel\": \"Cancel\",\n \"generating\": \"Generating...\",\n \"failed\": \"Failed\",\n \"lipSync\": \"Lip Sync\",\n \"lipSyncVideo\": \"Lip Sync Video\",\n \"lipSyncLabel\": \"Lip Sync\",\n \"lipSyncTitle\": \"Lip Sync\",\n \"original\": \"Original\",\n \"synced\": \"Synced\",\n \"videoFixed\": \"✓ Video\",\n \"imagePreview\": \"Image Preview\",\n \"playVoice\": \"Play Voice\",\n \"stopVoice\": \"Stop\",\n \"noVoice\": \"No voice available\",\n \"forceRegenerate\": \"Force Regenerate (use if stuck)\",\n \"regenerateVideo\": \"Regenerate Video\",\n \"lipSyncStatus\": \"Lip syncing...\",\n \"lipSyncInProgress\": \"Lip syncing in progress...\",\n \"lipSyncMayTakeMinutes\": \"This may take a few minutes\",\n \"audioEnabled\": \"Audio enabled\",\n \"audioDisabled\": \"Audio disabled\",\n \"isSynced\": \"(Synced)\",\n \"needVideo\": \"(Please generate video first)\",\n \"needAudio\": \"(Please generate audio first)\",\n \"generateAudio\": \"Generate Audio\",\n \"regenerateLipSync\": \"Regenerate Lip Sync\",\n \"editPrompt\": \"Edit Prompt\",\n \"clickToEditPrompt\": \"Click to edit prompt...\",\n \"shot\": \"Shot {number}\",\n \"unknownShotType\": \"Unknown\",\n \"correspondingText\": \"Corresponding Text\",\n \"generateVideo\": \"Generate Video\",\n \"selectModel\": \"Select Video Model\",\n \"selectVoice\": \"Select voice to use:\",\n \"willAutoPad\": \"(will auto-pad)\",\n \"autoPadding\": \"Padding\",\n \"redo\": \"Redo\",\n \"generatingAudio\": \"Generating...\",\n \"error\": {\n \"audioFailed\": \"Audio generation failed\"\n },\n \"batchMode\": \"Batch Mode\",\n \"batchModeDesc\": \"Offline inference, 50% cheaper, completes within 24 hours\",\n \"batchModeEnabled\": \"Batch mode enabled\",\n \"batchModeDisabled\": \"Batch mode disabled\"\n },\n \"promptModal\": {\n \"title\": \"Edit Shot #{number} Video Prompt\",\n \"shotType\": \"Shot Type:\",\n \"duration\": \"s\",\n \"location\": \"Location:\",\n \"locationUnknown\": \"Unknown\",\n \"characters\": \"Characters:\",\n \"charactersNone\": \"None\",\n \"description\": \"Description:\",\n \"text\": \"Corresponding Text:\",\n \"promptLabel\": \"Video Prompt\",\n \"placeholder\": \"Enter video prompt...\",\n \"tip\": \"Tip: Video models don't recognize character names, use appearance descriptions like \\\"young man with black hair and blue eyes\\\" instead of \\\"Victor\\\"\",\n \"save\": \"Save\",\n \"cancel\": \"Cancel\"\n },\n \"toolbar\": {\n \"title\": \"Video Generation\",\n \"filter\": \"Filter\",\n \"viewAll\": \"View All\",\n \"showGenerated\": \"Generated\",\n \"showPending\": \"Pending\",\n \"showFailed\": \"Failed\",\n \"totalShots\": \"Total {count} shots\",\n \"generatingShots\": \"{count} generating\",\n \"completedShots\": \"{count} completed\",\n \"failedShots\": \"{count} failed\",\n \"generateAll\": \"Generate All Videos\",\n \"batchConfigTitle\": \"Batch Generation Settings\",\n \"batchConfigDesc\": \"Choose model and parameters before generating all videos.\",\n \"confirmGenerateAll\": \"Confirm and Generate All\",\n \"confirming\": \"Submitting...\",\n \"noVideos\": \"No videos to download\",\n \"downloadCount\": \"Download {count} videos\",\n \"packing\": \"Packing...\",\n \"downloadAll\": \"Download All\",\n \"enterEditor\": \"Enter Video Editor\",\n \"enterEdit\": \"Enter Editor\",\n \"back\": \"Back\"\n },\n \"stage\": {\n \"title\": \"Video Generation\",\n \"generateAll\": \"Generate All\",\n \"regenerateFailed\": \"Retry Failed\",\n \"downloadAll\": \"Download All Videos\",\n \"enterEditor\": \"Enter Editor\",\n \"lipSyncStatus\": \"Lip syncing...\",\n \"hasSynced\": \"✓ Generated\",\n \"generating\": \"Generating...\",\n \"downloading\": \"Downloading...\",\n \"downloadProgress\": \"Preparing video files... {current}/{total}\",\n \"noVideos\": \"No generated videos\",\n \"scrollTo\": \"Jump to shot\",\n \"error\": {\n \"saveFailed\": \"Failed to save video prompt\",\n \"lipSyncFailed\": \"Lip sync failed\",\n \"fetchVideosFailed\": \"Failed to fetch video list\"\n },\n \"downloadFailed\": \"Download failed\",\n \"unknownError\": \"Unknown error\"\n },\n \"firstLastFrame\": {\n \"title\": \"First/Last Frame Settings\",\n \"firstFrame\": \"First Frame\",\n \"lastFrame\": \"Last Frame\",\n \"range\": \"Shot {from} → Shot {to}\",\n \"link\": \"Link\",\n \"unlink\": \"Unlink\",\n \"unlinkAction\": \"Unlink\",\n \"asLastFrameFor\": \"As last frame for Shot {number}\",\n \"asFirstFrameFor\": \"As first frame for Shot {number}\",\n \"customPrompt\": \"Custom Prompt\",\n \"promptPlaceholder\": \"Enter first/last frame video prompt...\",\n \"useDefault\": \"Use Default\",\n \"generate\": \"Generate First/Last Frame Video\",\n \"generated\": \"First/Last Frame Video Generated\",\n \"model\": \"Model\",\n \"withAudio\": \"With Audio\",\n \"audioOn\": \"On\",\n \"audioOff\": \"Off\",\n \"linkToNext\": \"Link to next panel (First/Last Frame)\",\n \"asLastFrame\": \"As last frame for Panel {number}\",\n \"thenTransitionTo\": \"then transition to\"\n },\n \"editor\": {\n \"alert\": {\n \"saveSuccess\": \"Saved successfully\",\n \"saveFailed\": \"Save failed\",\n \"exportStarted\": \"Export has started. Please wait...\",\n \"exportFailed\": \"Export failed\"\n },\n \"toolbar\": {\n \"back\": \"← Back\",\n \"saveDirty\": \"Save *\",\n \"saved\": \"Saved\",\n \"export\": \"Export Video\"\n },\n \"left\": {\n \"title\": \"Media Library\",\n \"description\": \"Clips imported from the video stage will appear here\"\n },\n \"right\": {\n \"title\": \"Properties\",\n \"clipLabel\": \"Clip:\",\n \"clipFallback\": \"Clip {index}\",\n \"durationLabel\": \"Duration:\",\n \"transitionLabel\": \"Transition to Next Clip\",\n \"deleteConfirm\": \"Delete this clip?\",\n \"deleteClip\": \"Delete Clip\",\n \"selectClipHint\": \"Select a clip to view properties\"\n },\n \"preview\": {\n \"emptyStartEditing\": \"Add media to start editing\"\n },\n \"timeline\": {\n \"zoomLabel\": \"Zoom:\",\n \"videoTrack\": \"Video\",\n \"emptyHint\": \"Drag video clips from media library to here\",\n \"audioTrack\": \"Voice\",\n \"audioBadge\": \"A\"\n },\n \"transition\": {\n \"title\": \"Transition\",\n \"duration\": \"Duration\",\n \"options\": {\n \"none\": \"None\",\n \"dissolve\": \"Dissolve\",\n \"fade\": \"Fade\",\n \"slide\": \"Slide\"\n }\n }\n },\n \"errors\": {\n \"unknownError\": \"Unknown error\"\n },\n \"capability\": {\n \"generationMode\": \"Generation mode\",\n \"generateAudio\": \"Generate audio\",\n \"duration\": \"Duration\",\n \"fps\": \"Frame rate\",\n \"resolution\": \"Resolution\",\n \"aspectRatio\": \"Aspect ratio\",\n \"reasoningEffort\": \"Reasoning effort\",\n \"voice\": \"Voice\",\n \"rate\": \"Rate\",\n \"mode\": \"Mode\"\n },\n \"unit\": {\n \"second\": \"s\",\n \"frame\": \"fps\"\n },\n \"common\": {\n \"generate\": \"Generate\"\n }\n}"
},
{
"path": "messages/en/voice.json",
"content": "{\n \"title\": \"Voice Lines\",\n \"linesCount\": \"{count} lines total, \",\n \"audioGeneratedCount\": \"{count} audio generated\",\n \"emotionPrompt\": \"Emotion Prompt\",\n \"emotionPromptTip\": \"(Leave empty for auto-reference)\",\n \"emotionPlaceholder\": \"e.g. laugh, English only...\",\n \"emotionStrength\": \"Emotion Strength\",\n \"flat\": \"Flat\",\n \"intense\": \"Intense\",\n \"generating\": \"Generating...\",\n \"generateVoice\": \"Generate Voice\",\n \"toolbar\": {\n \"back\": \"← Back\",\n \"analyzeLines\": \"Analyze Lines\",\n \"addLine\": \"+ Add Voice\",\n \"generateAll\": \"Generate All Voices\",\n \"downloadAll\": \"Download Voices\",\n \"generatingCount\": \"Generating ({count})\",\n \"packing\": \"Packing...\",\n \"stats\": \"{total} lines | {withVoice} with voice | {withAudio} generated\",\n \"noDownload\": \"No voices to download\",\n \"downloadCount\": \"Download {count} voices\",\n \"uploadReferenceHint\": \"Please upload reference audio for all characters first\"\n },\n \"speakerVoice\": {\n \"title\": \"Speaker Voice Status\",\n \"hint\": \"Please upload reference audio for characters in Asset Library\",\n \"linesCount\": \"{count} lines\",\n \"noVoice\": \"No reference voice\",\n \"configured\": \"✓ Configured\",\n \"playVoice\": \"Play current voice\",\n \"aiDesign\": \"AI Design Voice\",\n \"aiDesignVoice\": \"AI Design Voice\",\n \"redesign\": \"Redesign voice with AI\",\n \"uploadAudio\": \"Upload audio\",\n \"uploading\": \"Uploading\",\n \"upload\": \"Upload\",\n \"microsoftVoice\": \"Microsoft Voice\",\n \"microsoft\": \"MS\",\n \"maleVoices\": \"Male\",\n \"femaleVoices\": \"Female\",\n \"openAssetLibrary\": \"Asset Library\",\n \"configuredStatus\": \"Voice Set\",\n \"pendingStatus\": \"Voice Pending\",\n \"voiceSettings\": \"Voice Settings\",\n \"inlineLabel\": \"Inline\"\n },\n \"inlineBinding\": {\n \"title\": \"Set voice for \\\"{speaker}\\\"\",\n \"description\": \"This speaker is not in the asset library. Choose a method to set a reference voice.\",\n \"selectFromLibrary\": \"Select from Voice Library\",\n \"selectFromLibraryDesc\": \"Choose an existing global voice\",\n \"uploadAudio\": \"Upload Reference Audio\",\n \"uploadAudioDesc\": \"Upload MP3, WAV or other audio files as reference voice\",\n \"uploadQwenHint\": \"Uploaded voices can only be synthesized with IndexTTS, not QwenTTS. QwenTTS requires an AI-designed voice.\",\n \"aiDesign\": \"AI Design Voice\",\n \"aiDesignDesc\": \"Use AI to generate an exclusive reference voice\"\n },\n \"embedded\": {\n \"linesStats\": \"{total} lines · {audio} generated\",\n \"reanalyze\": \"Re-analyze\",\n \"analyzeLines\": \"Analyze Lines\",\n \"reanalyzeHint\": \"Re-analyze lines and update shot matching\",\n \"analyzeHint\": \"Extract lines from script\",\n \"downloadVoice\": \"Download Voices\",\n \"generateAllVoice\": \"Generate All Voices\",\n \"pendingCount\": \"({count} pending)\",\n \"generatingProgress\": \"Generating ({current}/{total})\",\n \"generatingHint\": \"Generating...\",\n \"noVoiceHint\": \"Please set voice for all characters above first\",\n \"noLinesHint\": \"No lines to generate\",\n \"allDoneHint\": \"All lines generated\",\n \"generateHint\": \"Click to generate {count} pending voices\",\n \"addLine\": \"+ Add Voice\",\n \"speakerVoiceStatus\": \"Speaker Voice Status\",\n \"speakersCount\": \"{count}\",\n \"listen\": \"Listen\",\n \"listenVoice\": \"Listen to voice\",\n \"reset\": \"Reset\",\n \"resetDesign\": \"Redesign\",\n \"aiDesign\": \"AI Design\",\n \"assetLibrary\": \"Asset Library\"\n },\n \"lineCard\": {\n \"generatingVoice\": \"Generating\",\n \"speaker\": \"Speaker\",\n \"speakerPlaceholder\": \"Speaker name\",\n \"content\": \"Content\",\n \"contentPlaceholder\": \"Line content\",\n \"emotionConfigured\": \"Emotion set\",\n \"emotionSettings\": \"Emotion Settings\",\n \"voiceConfigured\": \"✓ Configured\",\n \"needVoice\": \"Please set voice above\",\n \"locatePanel\": \"Locate bound shot\",\n \"locateVideo\": \"Locate Video\",\n \"play\": \"Play\",\n \"pause\": \"Pause\",\n \"locatePanelCta\": \"Jump to Shot {index}\",\n \"editLine\": \"Edit line\",\n \"deleteLine\": \"Delete line\",\n \"deleteAudio\": \"Delete audio\"\n },\n \"lineEditor\": {\n \"addTitle\": \"Add Voice Line\",\n \"editTitle\": \"Edit Voice Line\",\n \"contentLabel\": \"Line Content\",\n \"contentPlaceholder\": \"Enter line content\",\n \"speakerLabel\": \"Speaker\",\n \"speakerPlaceholder\": \"Enter speaker name\",\n \"selectSpeaker\": \"Select a speaker\",\n \"noSpeakerOptions\": \"No available speakers in this project yet. Analyze lines first.\",\n \"bindPanelLabel\": \"Bind Shot\",\n \"unboundPanel\": \"Unbound\",\n \"panelLabel\": \"Shot {index}\",\n \"saveAdd\": \"Add Voice\",\n \"saveEdit\": \"Save Changes\"\n },\n \"empty\": {\n \"title\": \"No Voice Lines\",\n \"description\": \"Extract lines and speakers from script\",\n \"analyzeButton\": \"Analyze Lines\",\n \"hint\": \"Please upload reference audio for characters in Asset Library first\"\n },\n \"confirm\": {\n \"deleteLine\": \"Are you sure you want to delete this line?\\n\\n\\\"{content}\\\"\\n\\nThis action cannot be undone.\",\n \"deleteAudio\": \"Are you sure you want to delete this audio?\\n\\n\\\"{content}\\\"\\n\\nThis action cannot be undone.\"\n },\n \"errors\": {\n \"saveFailed\": \"Save Failed\",\n \"analyzeFailed\": \"Analysis Failed\",\n \"generateFailed\": \"Generation Failed\",\n \"batchFailed\": \"Batch Generation Failed\",\n \"downloadFailed\": \"Download Failed\",\n \"deleteFailed\": \"Delete Failed\",\n \"addFailed\": \"Add Voice Failed\",\n \"invalidLineInput\": \"Content and speaker cannot be empty\",\n \"bindFailed\": \"Bind shot failed\",\n \"deleteAudioFailed\": \"Delete Audio Failed\",\n \"uploadFailed\": \"Upload Failed\",\n \"voiceDesignFailed\": \"Voice Design Save Failed\",\n \"emotionSaveFailed\": \"Emotion Settings Save Failed\",\n \"voiceGenerateFailed\": \"Voice Generation Failed\"\n },\n \"alerts\": {\n \"insufficientBalance\": \"Insufficient Balance\",\n \"insufficientBalanceMsg\": \"Account balance insufficient, please top up to continue\",\n \"noLinesToGenerate\": \"No lines to generate (please upload reference audio for characters first)\",\n \"generateComplete\": \"Complete: {success}/{total} successful\",\n \"generateFailed\": \"{count} failed\",\n \"speakerVoiceSet\": \"Reference audio set for {speaker}\",\n \"speakerVoiceUploaded\": \"Reference audio uploaded for {speaker}\",\n \"voiceDesignSet\": \"AI designed voice set for {speaker}\"\n },\n \"common\": {\n \"loading\": \"Loading...\",\n \"save\": \"Save\",\n \"cancel\": \"Cancel\",\n \"cancelling\": \"Cancelling...\",\n \"upload\": \"Upload\",\n \"download\": \"Download\",\n \"generate\": \"Generate\",\n \"regenerate\": \"Regenerate\"\n },\n \"assets\": {\n \"image\": {\n \"uploadFailed\": \"Upload Failed\"\n },\n \"stage\": {\n \"analyzing\": \"Analyzing...\"\n }\n },\n \"smartImport\": {\n \"errors\": {\n \"analyzeFailed\": \"Analysis Failed\"\n }\n },\n \"video\": {\n \"panelCard\": {\n \"play\": \"Play\"\n }\n },\n \"tts\": {\n \"generatedAudio\": \"Generated Audio\",\n \"browserNotSupport\": \"Your browser does not support audio playback\",\n \"audioDuration\": \"Audio Duration:\",\n \"subtitleCount\": \"Subtitle Count:\",\n \"noAudio\": \"No audio\",\n \"srtPreview\": \"SRT Subtitle Preview\",\n \"noSubtitle\": \"No subtitles\",\n \"stats\": \"Generation Stats\",\n \"minute\": \"min\",\n \"second\": \"sec\",\n \"items\": \"items\",\n \"completed\": \"✓ Completed\",\n \"regenerating\": \"Regenerating...\",\n \"regenerateTTS\": \"Regenerate TTS\",\n \"nextStep\": \"Next: Analyze Assets\",\n \"readyTip\": \"Click to proceed to asset analysis\",\n \"needGenerate\": \"Please generate TTS audio first\"\n },\n \"voiceCreate\": {\n \"aiDesignMode\": \"AI Design Voice\",\n \"uploadMode\": \"Upload Audio\",\n \"dropOrClick\": \"Drop file or click to select\",\n \"supportedFormats\": \"Supports MP3, WAV, OGG, M4A, AAC formats\",\n \"invalidFileType\": \"Unsupported file format. Please upload an audio file\",\n \"fileTooLarge\": \"File too large. Maximum 50MB supported\",\n \"previewAudio\": \"Preview Audio\",\n \"uploading\": \"Uploading...\",\n \"uploadFailed\": \"Upload failed\",\n \"uploadSuccess\": \"Upload successful\"\n },\n \"voiceDesign\": {\n \"presets\": {\n \"maleBroadcaster\": \"Male Broadcaster\",\n \"gentleFemale\": \"Gentle Female\",\n \"matureMale\": \"Mature Male\",\n \"livelyFemale\": \"Lively Girl\",\n \"intellectualFemale\": \"Intellectual Female\",\n \"narrator\": \"Narrator\"\n },\n \"presetsPrompts\": {\n \"maleBroadcaster\": \"Middle-aged male broadcaster with steady, deep voice, clear pronunciation\",\n \"gentleFemale\": \"Gentle and sweet young woman with clear, melodious voice\",\n \"matureMale\": \"Mature male with charismatic and expressive voice\",\n \"livelyFemale\": \"Lively young girl with sweet, cute, energetic voice\",\n \"intellectualFemale\": \"Elegant intellectual woman with clear, pleasant voice\",\n \"narrator\": \"Emotional narrator with warm, storytelling voice\"\n },\n \"defaultPreviewText\": \"Hello, nice to meet you. This is your AI-designed exclusive voice. Let me demonstrate its features for you. Whether it's a gentle conversation or an excited narration, I can deliver it perfectly. I hope you enjoy this voice, let's create amazing content together.\",\n \"pleaseSelectStyle\": \"Please enter or select a voice style\",\n \"designVoiceFor\": \"Design AI Voice for \\\"{speaker}\\\"\",\n \"hasExistingVoice\": \"Has voice\",\n \"selectStyle\": \"Select voice style:\",\n \"orCustomDescription\": \"Or custom description:\",\n \"describePlaceholder\": \"Describe voice characteristics: age, gender, tone, pitch...\",\n \"generateSchemesPrefix\": \"Generate\",\n \"generateSchemesSuffix\": \"Voice Schemes\",\n \"schemeCountAriaLabel\": \"Number of voice schemes to generate\",\n \"editPreviewText\": \"Edit preview text\",\n \"generating3Schemes\": \"Generating 3 voice schemes...\",\n \"estimatedTime\": \"Est. 15-30 seconds\",\n \"selectScheme\": \"Select voice scheme:\",\n \"schemeN\": \"Scheme {n}\",\n \"regenerate\": \"Regenerate\",\n \"confirmUse\": \"✓ Confirm Use\",\n \"confirmReplace\": \"Confirm Replace Voice?\",\n \"replaceWarning\": \"'s original voice, irreversible\",\n \"confirmReplaceBtn\": \"Confirm Replace\",\n \"noVoiceGenerated\": \"No voice generated\",\n \"generationError\": \"Voice generation failed\",\n \"generateFailed\": \"Failed to generate voice {n}\",\n \"preview\": \"Preview\",\n \"playing\": \"Playing\"\n }\n}\n"
},
{
"path": "messages/en/workspace.json",
"content": "{\n \"title\": \"My Projects\",\n \"subtitle\": \"Manage your AI anime production projects\",\n \"newProject\": \"New Project\",\n \"searchPlaceholder\": \"Search project name or description...\",\n \"searchButton\": \"Search\",\n \"clearButton\": \"Clear\",\n \"updatedAt\": \"Updated at\",\n \"noProjects\": \"No projects yet\",\n \"noProjectsDesc\": \"Create your first AI anime production project\",\n \"noResults\": \"No matching projects found\",\n \"noResultsDesc\": \"Try using different search terms\",\n \"createProject\": \"New Project\",\n \"editProject\": \"Edit Project\",\n \"deleteProject\": \"Delete Project\",\n \"deleteConfirm\": \"Are you sure you want to delete project \\\"{name}\\\"? This action cannot be undone.\",\n \"projectName\": \"Project Name\",\n \"projectNamePlaceholder\": \"Enter project name\",\n \"projectDescription\": \"Project Description (Optional)\",\n \"projectDescriptionPlaceholder\": \"Enter project description\",\n \"creating\": \"Creating...\",\n \"saving\": \"Saving...\",\n \"createFailed\": \"Failed to create project\",\n \"analysisModelRequiredAfterCreate\": \"Project created. Please configure default models in Profile first (at minimum an analysis model), or the project cannot be used.\",\n \"updateFailed\": \"Failed to update project\",\n \"deleteFailed\": \"Failed to delete project\",\n \"totalProjects\": \"{count} projects in total\",\n \"statsEpisodes\": \"Episodes\",\n \"statsImages\": \"Images\",\n \"statsVideos\": \"Videos\",\n \"noContent\": \"No content yet\",\n \"modelNotConfigured\": {\n \"before\": \"No models configured yet. Please go to\",\n \"link\": \"Settings Center\",\n \"after\": \"to configure models, or customize them in project settings after creation.\"\n }\n}"
},
{
"path": "messages/en/workspaceDetail.json",
"content": "{\n \"globalAssets\": \"Global Assets\",\n \"createFailed\": \"Creation failed\",\n \"deleteFailed\": \"Deletion failed\",\n \"renameFailed\": \"Rename failed\",\n \"refreshFailed\": \"Refresh failed\",\n \"projectNotFound\": \"Project not found\",\n \"backToWorkspace\": \"Back to Workspace\",\n \"episode\": \"Episode\",\n \"modelSetup\": {\n \"title\": \"Set up default models before starting\",\n \"description\": \"No default analysis model is configured for this account. Smart split and downstream AI flows cannot run until it is set.\",\n \"configureNow\": \"Configure Analysis Model\",\n \"goProfile\": \"Go to Profile Settings\",\n \"modalTitle\": \"Configure Default Analysis Model\",\n \"modalDescription\": \"Select an available text model as the default analysis model.\",\n \"selectModelLabel\": \"Default Analysis Model\",\n \"selectModelPlaceholder\": \"Select an analysis model\",\n \"selectModelFirst\": \"Please select an analysis model first\",\n \"noModelOptions\": \"No text model is available. Configure your provider and models in Profile settings first.\",\n \"saveFailed\": \"Failed to save default analysis model. Please try again.\"\n },\n \"sidebar\": {\n \"dragToMove\": \"Drag to move position\",\n \"listTitle\": \"Episode List\",\n \"episodeCount\": \"{count} episodes\",\n \"empty\": \"No episodes yet. Create one below.\",\n \"save\": \"Save\",\n \"deleteConfirm\": \"Delete \\\"{name}\\\"?\",\n \"delete\": \"Delete\",\n \"cancel\": \"Cancel\",\n \"rename\": \"Rename\",\n \"newEpisodePlaceholder\": \"Enter episode name...\",\n \"create\": \"Create\",\n \"addEpisode\": \"Add Episode\"\n }\n}\n"
},
{
"path": "messages/en/worldContextModal.json",
"content": "{\n \"title\": \"World & Character Settings\",\n \"description\": \"Define global character appearances, scene styles, and environment descriptions\",\n \"placeholder\": \"Example:\\n【Protagonist】John, 25, short black hair, always wearing a faded denim jacket, melancholy eyes.\\n【Heroine】Jane, 22, red twin tails, lively personality, likes to wear Lolita-style dresses.\\n【Scene】Cyberpunk city in 2077, neon lights flashing, raining all year round...\",\n \"hint\": \"These settings will be inherited by all episodes as a reference for AI drawing.\"\n}"
},
{
"path": "messages/zh/actions.json",
"content": "{\n \"storyboard\": \"分镜图\",\n \"storyboard_candidate\": \"分镜候选\",\n \"character\": \"角色图\",\n \"location\": \"场景图\",\n \"video\": \"视频\",\n \"analyze\": \"分析\",\n \"analyze_character\": \"角色分析\",\n \"analyze_location\": \"场景分析\",\n \"clips\": \"片段切割\",\n \"storyboard_text_plan\": \"分镜规划\",\n \"storyboard_text_detail\": \"分镜细节\",\n \"tts\": \"语音合成\",\n \"regenerate\": \"重生成\",\n \"voice-generate\": \"配音生成\",\n \"voice-design\": \"声音设计\",\n \"lip-sync\": \"口型同步\"\n}"
},
{
"path": "messages/zh/apiConfig.json",
"content": "{\n \"title\": \"API 配置\",\n \"saving\": \"保存中...\",\n \"saved\": \"已保存\",\n \"saveFailed\": \"保存失败\",\n \"connected\": \"已连接\",\n \"notConfigured\": \"未配置 Key\",\n \"configure\": \"配置\",\n \"connect\": \"连接\",\n \"compatibilityLayerOpenAI\": \"OpenAI 兼容层\",\n \"compatibilityLayerGemini\": \"Gemini 兼容层\",\n \"show\": \"显示\",\n \"hide\": \"隐藏\",\n \"capability\": \"能力\",\n \"default\": \"默认\",\n \"delete\": \"删除\",\n \"add\": \"添加\",\n \"cancel\": \"取消\",\n \"close\": \"关闭\",\n \"save\": \"保存\",\n \"comingSoon\": \"待上线\",\n \"priceInput\": \"输入 {amount}\",\n \"priceOutput\": \"输出 {amount}\",\n \"priceUnavailable\": \"暂无定价\",\n \"fillComplete\": \"请填写完整信息\",\n \"fillPricing\": \"请填写价格信息\",\n \"pricingInputLabel\": \"输入价格\",\n \"pricingOutputLabel\": \"输出价格\",\n \"pricingBasePriceLabel\": \"基础价格（可选）\",\n \"pricingEnableCustom\": \"启用自定义价格（可选）\",\n \"pricingOptionPricesPlaceholder\": \"参数价格 JSON（可选），示例：{\\\"resolution\\\":{\\\"1024x1024\\\":0.12}}\",\n \"modelIdExists\": \"模型 ID 已存在\",\n \"flushConfigFailed\": \"保存厂商配置失败，请先检查并保存 API Key / Base URL\",\n \"probeLlmProtocolFailed\": \"模型协议探测失败，请稍后重试\",\n \"probeAuthFailed\": \"模型协议探测鉴权失败，请检查 API Key\",\n \"probeInconclusive\": \"模型协议探测结果不确定（可能限流或服务异常），请稍后重试\",\n \"probeRequestFailed\": \"模型协议探测请求失败，请稍后重试\",\n \"modelDisplayName\": \"外显名称（你自己看的）\",\n \"modelActualId\": \"实际调用 ID（模型参数）\",\n \"noModelsForProvider\": \"该厂商暂无配置模型\",\n \"defaultModels\": \"默认模型配置\",\n \"textDefault\": \"文本模型\",\n \"characterDefault\": \"角色模型\",\n \"locationDefault\": \"场景模型\",\n \"storyboardDefault\": \"分镜模型\",\n \"editDefault\": \"修图模型\",\n \"videoDefault\": \"视频模型\",\n \"audioDefault\": \"语音模型\",\n \"lipsyncDefault\": \"口型同步模型\",\n \"selectDefault\": \"请选择\",\n \"defaultModelDesc\": {\n \"analysisModel\": \"负责剧本解析、分镜构建等全流程文本分析能力。\",\n \"videoModel\": \"负责将图像与指令合成为最终视频片段。\",\n \"characterModel\": \"根据剧本描述生成角色形象与外观参考图\",\n \"locationModel\": \"根据剧本描述生成场景环境与空间参考图\",\n \"storyboardModel\": \"根据分镜脚本生成对应的镜头画面参考图\",\n \"editModel\": \"对已有图片进行局部修改、风格调整与精修\",\n \"audioModel\": \"将文本台词转化为自然流畅的语音音频\",\n \"lipSyncModel\": \"将语音与视频角色的口型进行精准同步\",\n \"voiceDesignModel\": \"为角色定制专属音色与语音风格特征\"\n },\n \"defaultModelSection\": {\n \"coreFoundation\": \"文本分析与视频能力\",\n \"creativePipeline\": \"全局图像模型配置\",\n \"unifiedOverride\": \"批量配置图像模型\",\n \"unifiedOverrideHint\": \"设置负责整个系统所有地方图像生成/编辑的模型\",\n \"unifiedOverridePlaceholder\": \"批量应用到以下场景...\",\n \"followUnified\": \"跟随全局配置\",\n \"extensions\": \"扩展功能\",\n \"coreTextTitle\": \"文本分析模型\",\n \"coreVideoTitle\": \"视频生成模型\",\n \"pipelineCharacter\": \"角色生成\",\n \"pipelineLocation\": \"场景生成\",\n \"pipelineStoryboard\": \"镜头生成\",\n \"pipelineEdit\": \"编辑图片\",\n \"extLipSync\": \"口型同步\",\n \"extTTS\": \"语音合成\",\n \"extVoiceDesign\": \"音色设计\",\n \"corePlaceholder\": \"必选配置\",\n \"extPlaceholder\": \"暂不启用\"\n },\n \"imageModelTip\": \"推荐使用 Google Banana 系列模型，目前其他图像模型生成效果有限。\",\n \"customProviderTip\": \"项目目前为测试版，由于市面上各厂商自定义 API 格式差异较大，自定义 API 兼容性尚不完善，建议优先使用官方内置 API。后续版本将持续更新以兼容更多厂商。\",\n \"providerPool\": \"厂商资源池\",\n \"providerPoolDesc\": \"在此使用来自全球丰富的模型配置\",\n \"dragToSort\": \"拖拽排序\",\n \"dragToSortHint\": \"按住左上角拖拽手柄可调整厂商顺序\",\n \"hideProvider\": \"隐藏提供商\",\n \"hideProviderConfirm\": \"确定要隐藏此提供商吗？隐藏后将置底显示，可随时恢复。\",\n \"showProvider\": \"显示提供商\",\n \"showHiddenProviders\": \"显示隐藏的提供商\",\n \"hideHiddenProviders\": \"收起隐藏的提供商\",\n \"hiddenProvidersPrefix\": \"已隐藏\",\n \"providerIdExists\": \"该提供商 ID 已存在\",\n \"presetProviderCannotDelete\": \"预设提供商不能删除\",\n \"confirmDeleteProvider\": \"确定删除这个提供商吗？\",\n \"presetModelCannotDelete\": \"预设模型不能删除\",\n \"confirmDeleteModel\": \"确定删除这个模型吗？\",\n \"addGeminiProvider\": \"新增模型服务商\",\n \"baseUrl\": \"Base URL\",\n \"configureBaseUrl\": \"配置地址\",\n \"addModel\": \"添加模型\",\n \"batchModeHalfPrice\": \"批量模式（价格减半）\",\n \"openaiCompatVideoOnlyHint\": \"仅支持 OpenAI 官方格式的图生视频模型。\",\n \"typeText\": \"文本\",\n \"typeImage\": \"图像\",\n \"typeVideo\": \"视频\",\n \"typeAudio\": \"音频等\",\n \"apiKeyLabel\": \"API Key\",\n \"apiType\": \"API 类型\",\n \"apiTypeGeminiCompatible\": \"Gemini 兼容\",\n \"apiTypeOpenAICompatible\": \"OpenAI 兼容\",\n \"apiTypeGeminiHint\": \"使用 Google SDK\",\n \"otherProviders\": \"其他配置\",\n \"audioCategory\": \"音频\",\n \"audioAndLipsync\": \"语音与唇形同步\",\n \"configureApiKey\": \"配置 API Key\",\n \"enterApiKey\": \"请输入 API Key\",\n \"testConnection\": \"测试连接\",\n \"testing\": \"正在测试...\",\n \"testPassed\": \"连接测试通过\",\n \"testFailed\": \"连接测试未通过\",\n \"testWarning\": \"建议检查配置后再添加\",\n \"testRetry\": \"重试\",\n \"addAnyway\": \"仍然添加\",\n \"testStep\": {\n \"models\": \"模型列表\",\n \"textGen\": \"文本生成\",\n \"imageGen\": \"图像生成\",\n \"credits\": \"余额检查\",\n \"audioGen\": \"音频生成\",\n \"skipped\": \"已跳过\"\n },\n \"tabs\": {\n \"llm\": \"文本模型\",\n \"image\": \"图片模型\",\n \"video\": \"视频模型\",\n \"audio\": \"音频等模型\",\n \"other\": \"其他\"\n },\n \"sections\": {\n \"llmApiKeys\": \"文本模型 API Keys\",\n \"imageApiKeys\": \"图片模型 API Keys\",\n \"videoApiKeys\": \"视频模型 API Keys\",\n \"audioApiKey\": \"音频模型 API Key\",\n \"lipsyncApiKey\": \"唇形同步 API Key\"\n },\n \"defaultModel\": {\n \"title\": \"默认模型\",\n \"hint\": \"新建项目与资产库将使用此默认配置，也可在项目设置中为单独项目自定义模型\",\n \"notSelected\": \"未选择\",\n \"analysis\": \"分析模型\",\n \"image\": \"图片生成\",\n \"video\": \"视频生成\",\n \"resolution\": \"图片分辨率\"\n },\n \"workflowConcurrency\": {\n \"analysis\": \"分析流程并发\",\n \"image\": \"图像流程并发\",\n \"video\": \"视频流程并发\"\n },\n \"viewTutorial\": \"查看教程\",\n \"tutorial\": {\n \"button\": \"开通教程\",\n \"title\": \"开通教程\",\n \"subtitle\": \"按照以下步骤完成配置\",\n \"close\": \"我知道了\",\n \"openLink\": \"点击打开\",\n \"steps\": {\n \"ark_step1\": \"进入火山引擎控制台，开通 API Key\",\n \"ark_step2\": \"在模型管理页面，点击右上角「一键开通所有模型」\",\n \"openrouter_step1\": \"进入 OpenRouter 平台，创建 API Key（必须选择带有图像功能的模型）\",\n \"fal_step1\": \"进入 FAL 平台，创建 API Key\",\n \"google_step1\": \"进入 Google AI Studio，创建 API Key\",\n \"minimax_step1\": \"进入海螺 MiniMax 平台，获取 API Key\",\n \"vidu_step1\": \"进入生数科技 Vidu 平台，点击「新建 API Key」\",\n \"openai_compatible_step1\": \"填写任意 OpenAI 协议兼容服务的 Base URL 与 API Key\",\n \"gemini_compatible_step1\": \"填写任意 Gemini 协议兼容服务的 Base URL 与 API Key\",\n \"bailian_step1\": \"进入阿里云百炼平台，获取 API Key\",\n \"siliconflow_step1\": \"进入硅基流动控制台，创建 API Key\"\n }\n },\n \"assistantOpen\": \"AI配置助手\",\n \"assistantTitle\": \"AI 配置助手\",\n \"assistantSubtitle\": \"通过对话把第三方文档转换为可执行图像/视频模板，并自动保存\",\n \"assistantWelcome\": \"请描述你的接口文档（端点、请求体、返回字段）。我会追问缺失信息并自动保存模型。\",\n \"assistantInputPlaceholder\": \"粘贴文档或描述接口细节...\",\n \"assistantSend\": \"发送\",\n \"assistantDisabledHint\": \"请先配置并保存 API Key 与 Base URL\",\n \"assistantRequestFailed\": \"助手请求失败，请稍后重试\",\n \"assistantResponseInvalid\": \"助手返回格式无效，请重试\",\n \"assistantMissingTitle\": \"待补充字段\",\n \"assistantWarningsTitle\": \"风险提示\",\n \"assistantDraftTitle\": \"当前草稿模型\",\n \"assistantReasoningTitle\": \"思考过程\",\n \"assistantReasoningExpand\": \"点击展开\",\n \"assistantReasoningCollapse\": \"点击收起\",\n \"assistantCompletedTitle\": \"模型模板已保存\",\n \"assistantCompletedMessage\": \"模型 {model} 已自动加入当前服务商。\\n请点击关闭结束本次对话。\",\n \"you\": \"你\",\n \"thinking\": \"思考中...\"\n}\n"
},
{
"path": "messages/zh/apiTypes.json",
"content": "{\n \"image\": \"图片生成\",\n \"video\": \"视频生成\",\n \"text\": \"文本分析\",\n \"tts\": \"语音合成\",\n \"voice\": \"配音\",\n \"voice_design\": \"声音设计\",\n \"lip_sync\": \"口型同步\"\n}"
},
{
"path": "messages/zh/assetHub.json",
"content": "{\n \"title\": \"资产中心\",\n \"description\": \"管理您的全局角色和场景资产\",\n \"modelHint\": \"资产中心使用默认模型，如需修改请在\",\n \"modelHintLink\": \"API配置页面\",\n \"modelHintSuffix\": \"进行设置\",\n \"folders\": \"文件夹\",\n \"noFolders\": \"暂无文件夹\",\n \"allAssets\": \"所有资产\",\n \"characters\": \"角色\",\n \"locations\": \"场景\",\n \"voices\": \"音色\",\n \"addCharacter\": \"新建角色\",\n \"addLocation\": \"新建场景\",\n \"addVoice\": \"新建音色\",\n \"downloadAll\": \"打包下载\",\n \"downloadAllTitle\": \"下载全部图片资产\",\n \"downloading\": \"打包中...\",\n \"downloadSuccess\": \"下载完成\",\n \"downloadFailed\": \"下载失败\",\n \"downloadEmpty\": \"当前没有可下载的图片资产\",\n \"newFolder\": \"新建文件夹\",\n \"editFolder\": \"编辑文件夹\",\n \"deleteFolder\": \"删除文件夹\",\n \"folderName\": \"文件夹名称\",\n \"folderNamePlaceholder\": \"请输入文件夹名称\",\n \"emptyState\": \"暂无资产\",\n \"emptyStateHint\": \"点击上方按钮添加角色或场景\",\n \"generate\": \"生成图片\",\n \"generating\": \"生成中...\",\n \"regenerate\": \"重新生成\",\n \"undo\": \"撤回\",\n \"delete\": \"删除\",\n \"cancel\": \"取消\",\n \"save\": \"保存\",\n \"create\": \"创建\",\n \"confirmDeleteFolder\": \"确定删除此文件夹吗？文件夹内的资产将移至未分类。\",\n \"confirmDeleteCharacter\": \"确定删除此角色吗？此操作无法撤回。\",\n \"confirmDeleteLocation\": \"确定删除此场景吗？此操作无法撤回。\",\n \"confirmDeleteVoice\": \"确定删除此音色吗？此操作无法撤回。\",\n \"voiceName\": \"音色名称\",\n \"voiceNamePlaceholder\": \"请输入音色名称\",\n \"voiceNameRequired\": \"请输入音色名称\",\n \"voicePickerTitle\": \"从音色库选择\",\n \"voicePickerEmpty\": \"暂无音色，请先创建音色\",\n \"voicePickerConfirm\": \"确认选择\",\n \"pagination\": {\n \"previous\": \"上一页\",\n \"next\": \"下一页\"\n },\n \"common\": {\n \"cancel\": \"取消\"\n },\n \"generateFailed\": \"生成失败\",\n \"selectFailed\": \"选择失败\",\n \"uploadFailed\": \"上传失败\",\n \"editFailed\": \"编辑失败\",\n \"saveVoiceFailed\": \"保存声音失败\",\n \"saveVoiceFailedDetail\": \"保存声音失败: {error}\",\n \"bindVoiceFailed\": \"绑定音色失败\",\n \"bindVoiceFailedDetail\": \"绑定音色失败: {error}\",\n \"voiceDesignSaved\": \"已为 {name} 设置 AI 设计的声音\",\n \"appearanceLabel\": \"形象 {index}\",\n \"voiceSettings\": {\n \"title\": \"配音音色\",\n \"noVoice\": \"无音色\",\n \"previewFailed\": \"预览失败: {error}\",\n \"uploadFailed\": \"上传音频失败: {error}\",\n \"uploading\": \"上传中...\",\n \"uploaded\": \"已上传\",\n \"uploadAudio\": \"上传音频\",\n \"aiDesign\": \"AI设计\",\n \"voiceLibrary\": \"音色库\",\n \"pause\": \"暂停\",\n \"preview\": \"试听音色\"\n },\n \"modal\": {\n \"newCharacter\": \"新建角色\",\n \"confirm\": \"确定\",\n \"processing\": \"处理中...\",\n \"newLocation\": \"新建场景\",\n \"addCharacter\": \"创建角色\",\n \"addLocation\": \"创建场景\",\n \"adding\": \"创建中...\",\n \"aiDesign\": \"AI 设计\",\n \"aiDesignPlaceholder\": \"例如：一个身穿红色长裙的古风美女，长发飘飘，手持折扇\",\n \"aiDesignLocationPlaceholder\": \"例如：一个古典中式园林，有假山流水和亭台楼阁\",\n \"aiDesignTip\": \"AI 将根据您的需求生成详细的描述，您可以在生成后编辑\",\n \"aiDesignLocationTip\": \"AI 将根据您的需求生成详细的场景描述\",\n \"generate\": \"生成\",\n \"generating\": \"生成中...\",\n \"nameLabel\": \"角色名称\",\n \"namePlaceholder\": \"请输入角色名称\",\n \"descLabel\": \"角色描述\",\n \"descPlaceholder\": \"请描述角色的外形特征、服装、发型等...\",\n \"locationNameLabel\": \"场景名称\",\n \"locationNamePlaceholder\": \"请输入场景名称\",\n \"locationSummaryLabel\": \"场景描述\",\n \"locationSummaryPlaceholder\": \"请描述场景的环境、氛围、特征等...\",\n \"referenceUpload\": \"上传参考图\",\n \"referenceUploadTip\": \"上传一张角色图片，AI 将自动转换为三视图设定图\",\n \"convertToCharacter\": \"转换三视图\",\n \"converting\": \"转换中...\",\n \"dropOrClick\": \"拖放图片或点击上传\",\n \"supportedFormats\": \"支持 JPG、PNG 格式\"\n }\n}"
},
{
"path": "messages/zh/assetLibrary.json",
"content": "{\n \"title\": \"资产库\",\n \"button\": \"资产库\",\n \"characters\": \"角色\",\n \"locations\": \"场景\",\n \"noCharacters\": \"暂无角色\",\n \"noLocations\": \"暂无场景\",\n \"addCharacter\": \"添加角色\",\n \"addLocation\": \"添加场景\",\n \"generateImage\": \"生成图片\",\n \"regenerateImage\": \"重新生成\",\n \"analyzeAssets\": \"分析资产\",\n \"analyzing\": \"分析中...\"\n}"
},
{
"path": "messages/zh/assetModal.json",
"content": "{\n \"character\": {\n \"title\": \"新建角色\",\n \"name\": \"角色名称\",\n \"namePlaceholder\": \"请输入角色名称\",\n \"modeReference\": \"参考图模式\",\n \"modeDescription\": \"描述模式\",\n \"isSubAppearance\": \"这是一个子形象\",\n \"isSubAppearanceHint\": \"为已有角色添加新的形象状态\",\n \"uploadReference\": \"上传参考图\",\n \"pasteHint\": \"Ctrl+V 粘贴\",\n \"dropOrClick\": \"点击上传或拖拽图片\",\n \"supportedFormats\": \"支持 JPG、PNG 等格式\",\n \"nameRequired\": \"请先输入角色名称才能使用参考图转换\",\n \"convertToSheet\": \"将参考图转换为标准三视图\",\n \"useReferenceGeneratePrefix\": \"使用参考图生成\",\n \"generateCountSuffix\": \"张图片\",\n \"selectReferenceGenerateCount\": \"选择参考图生成数量\",\n \"referenceTip\": \"上传任意角色图片，AI 将自动生成标准三视图设定\",\n \"description\": \"角色描述\",\n \"modifyDescription\": \"修改描述\",\n \"descPlaceholder\": \"请输入角色外貌描述...\",\n \"modifyDescriptionPlaceholder\": \"描述要对主形象做什么修改,例如:换上正装、受伤后的状态、披上斗篷...\",\n \"selectMainCharacter\": \"选择主角色\",\n \"selectCharacterPlaceholder\": \"请选择角色...\",\n \"appearancesCount\": \"{count} 个形象\",\n \"changeReason\": \"形象变化原因\",\n \"changeReasonPlaceholder\": \"例如:战斗后受伤、穿上正装参加宴会...\",\n \"defaultDescription\": \"{name} 的角色设定\",\n \"generationMode\": \"生成方式\",\n \"directGenerate\": \"直接生成\",\n \"extractPrompt\": \"反推提示词\",\n \"extractFirst\": \"先提取描述\",\n \"directGenerateDesc\": \"直接使用参考图生成角色设定图（图生图）\",\n \"extractPromptDesc\": \"先从图片提取角色描述，可编辑后再生成（文生图）\",\n \"maxReferenceImages\": \"最多上传 5 张参考图\",\n \"selectedCount\": \"已选择 {count}/5 张参考图\",\n \"extractDescription\": \"提取角色描述\",\n \"extracting\": \"提取中...\",\n \"extractedDescription\": \"提取的角色描述（可编辑）\",\n \"reExtract\": \"重新提取\",\n \"editHint\": \"编辑后点击下方按钮生成角色设定图\",\n \"generateFromDescription\": \"使用描述生成\",\n \"textToImageTip\": \"文生图模式：基于提取的描述生成角色设定图\",\n \"pleaseExtractFirst\": \"请先提取角色描述\"\n },\n \"location\": {\n \"title\": \"新建场景\",\n \"name\": \"场景名称\",\n \"namePlaceholder\": \"请输入场景名称\",\n \"description\": \"场景描述\",\n \"descPlaceholder\": \"请输入场景描述...\"\n },\n \"artStyle\": {\n \"title\": \"画面风格\"\n },\n \"aiDesign\": {\n \"title\": \"AI 设计\",\n \"placeholder\": \"描述你想要的角色特征...\",\n \"placeholderLocation\": \"描述场景氛围和环境...\",\n \"generating\": \"设计中...\",\n \"generate\": \"生成\",\n \"tip\": \"输入简单描述，AI 帮你生成详细设定\"\n },\n \"common\": {\n \"creating\": \"创建中...\",\n \"create\": \"创建\",\n \"cancel\": \"取消\",\n \"adding\": \"添加中...\",\n \"add\": \"添加\",\n \"addOnly\": \"仅添加角色\",\n \"addOnlyToAssetHub\": \"仅添加人物到资产库\",\n \"addOnlyLocation\": \"仅添加场景\",\n \"addOnlyToAssetHubLocation\": \"仅添加场景到资产库\",\n \"addAndGeneratePrefix\": \"添加并生成\",\n \"generateCountSuffix\": \"张图片\",\n \"selectGenerateCount\": \"选择生成数量\",\n \"optional\": \"（可选）\"\n },\n \"errors\": {\n \"uploadFailed\": \"上传失败\",\n \"extractDescriptionFailed\": \"提取描述失败\",\n \"createFailed\": \"创建失败\",\n \"aiDesignFailed\": \"AI 设计失败\",\n \"addSubAppearanceFailed\": \"添加子形象失败\",\n \"insufficientBalance\": \"账户余额不足\"\n }\n}\n"
},
{
"path": "messages/zh/assetPicker.json",
"content": "{\n \"selectCharacter\": \"从资产中心选择角色\",\n \"selectLocation\": \"从资产中心选择场景\",\n \"selectVoice\": \"从资产中心选择音色\",\n \"searchPlaceholder\": \"搜索资产名称或文件夹...\",\n \"noAssets\": \"资产中心暂无资产\",\n \"createInAssetHub\": \"请先在资产中心创建角色/场景/音色\",\n \"noSearchResults\": \"未找到匹配的资产\",\n \"appearances\": \"个形象\",\n \"images\": \"张图片\",\n \"cancel\": \"取消\",\n \"confirmCopy\": \"确认复制\",\n \"copyFromGlobal\": \"从资产中心复制\",\n \"copySuccess\": \"复制成功\",\n \"copyFailed\": \"复制失败\",\n \"preview\": \"试听\",\n \"stop\": \"停止\"\n}"
},
{
"path": "messages/zh/assets.json",
"content": "{\n \"stage\": {\n \"title\": \"资产确认\",\n \"characters\": \"角色\",\n \"locations\": \"场景\",\n \"analyze\": \"分析资产\",\n \"analyzing\": \"分析中...\",\n \"generateAll\": \"批量生成全部\",\n \"noCharacters\": \"暂无角色\",\n \"noLocations\": \"暂无场景\",\n \"confirmProfiles\": \"角色档案待确认\",\n \"confirmHint\": \"请确认以下角色档案后生成外貌描述\",\n \"confirmAll\": \"全部确认 ({count})\",\n \"assetsTitle\": \"资产分析\",\n \"characterAssets\": \"角色资产\",\n \"locationAssets\": \"场景资产\",\n \"counts\": \"{characterCount} 个角色，{appearanceCount} 个形象\",\n \"locationCounts\": \"{count} 个场景\",\n \"undoFailed\": \"撤回失败\",\n \"undoFailedError\": \"撤回失败: {error}\",\n \"undoSuccess\": \"已撤回到上一版本\",\n \"editFailed\": \"编辑失败\",\n \"editFailedError\": \"图片编辑失败: {error}\",\n \"updateSuccess\": \"描述词已同步更新\"\n },\n \"character\": {\n \"add\": \"添加角色\",\n \"edit\": \"编辑角色\",\n \"delete\": \"删除角色\",\n \"deleteConfirm\": \"确定要删除这个角色吗？\",\n \"deleteAppearanceConfirm\": \"确定要删除这个形象吗？\",\n \"deleteFailed\": \"删除失败: {error}\",\n \"deleteWhole\": \"删除整个角色\",\n \"deleteOptions\": \"删除选项\",\n \"name\": \"角色名\",\n \"description\": \"外貌描述\",\n \"generateImage\": \"生成形象\",\n \"regenerateImage\": \"重新生成\",\n \"generate\": \"生成\",\n \"regenerating\": \"生成中...\",\n \"profile\": \"设定档案\",\n \"voiceSettings\": \"配音设置\",\n \"speaker\": \"说话人\",\n \"selectSpeaker\": \"选择说话人\",\n \"noSpeaker\": \"未设置\",\n \"primary\": \"主形象\",\n \"secondary\": \"子状态\",\n \"generateFromPrimary\": \"基于主形象生成\",\n \"selectPrimaryFirst\": \"请先选择主形象\",\n \"editing\": \"编辑中...\",\n \"confirming\": \"确认中...\",\n \"assetCount\": \"{count} 个形象\",\n \"characterCount\": \"{count} 个角色\",\n \"updateFailed\": \"更新描述失败\",\n \"addFailed\": \"添加角色失败\",\n \"copyFromGlobal\": \"从资产中心复制\"\n },\n \"location\": {\n \"add\": \"添加场景\",\n \"edit\": \"编辑场景\",\n \"delete\": \"删除场景\",\n \"deleteConfirm\": \"确定要删除这个场景吗？\",\n \"deleteFailed\": \"删除失败: {error}\",\n \"name\": \"场景名\",\n \"summary\": \"简要描述\",\n \"summaryPlaceholder\": \"场景用途/人物关联，如：张三居住的主卧室\",\n \"description\": \"场景描述\",\n \"generateImage\": \"生成图片\",\n \"regenerateImage\": \"重新生成\",\n \"updateFailed\": \"更新描述失败\",\n \"addFailed\": \"添加场景失败\"\n },\n \"image\": {\n \"upload\": \"上传图片\",\n \"uploadReplace\": \"上传替换图片\",\n \"uploadFailed\": \"上传失败\",\n \"uploadFailedError\": \"上传失败: {error}\",\n \"uploadSuccess\": \"上传成功！\",\n \"edit\": \"编辑图片\",\n \"editPrompt\": \"编辑提示词\",\n \"undo\": \"撤回到上一版本\",\n \"undoConfirm\": \"确定要撤回到上一版本吗？当前版本将被删除。\",\n \"regenerateGroup\": \"整组重新生成\",\n \"regenerateStuck\": \"点击重新生成（当前卡住）\",\n \"selectCount\": \"选择生成数量\",\n \"generateCountPrefix\": \"生成\",\n \"generateCountSuffix\": \"张图像\",\n \"regenCountPrefix\": \"重新生成\",\n \"regenCountSuffix\": \"张\",\n \"regenCountAriaLabel\": \"选择重新生成张数\",\n \"generatedProgress\": \"已生成 {generated}/{total}\",\n \"generating\": \"生成中\",\n \"regenerating\": \"重新生成中\",\n \"generatingPlaceholder\": \"待生成\",\n \"selectTip\": \"选择并确认后，可对图片进行编辑和修改\",\n \"selectFirst\": \"请选择一张图片\",\n \"useThis\": \"选择此方案\",\n \"optionAlt\": \"{name} - 方案{number}\",\n \"optionNumber\": \"方案{number}\",\n \"optionSelected\": \"已选择方案{number}\",\n \"confirmOption\": \"确定选择方案{number}\",\n \"deleteOthersHint\": \"（删除其他）\",\n \"confirmSuccess\": \"确认成功\",\n \"confirmFailed\": \"确认选择失败: {error}\",\n \"selectFailed\": \"选择图片失败: {error}\",\n \"cancelSelection\": \"取消选择\",\n \"deleteThis\": \"删除此形象\",\n \"undoFailed\": \"撤回失败\",\n \"undoSuccess\": \"✓ 已撤回到上一版本\",\n \"editFailed\": \"图片编辑失败\",\n \"editSuccess\": \"图片编辑成功\",\n \"regenerateFailed\": \"重新生成失败: {error}\"\n },\n \"modal\": {\n \"newCharacter\": \"新增角色\",\n \"addSubAppearance\": \"添加子形象\",\n \"aiDesign\": \"AI智能设计\",\n \"aiDesigning\": \"AI设计中...\",\n \"designInstruction\": \"请输入设计指令\",\n \"enterNameDesc\": \"请填写角色名称和描述\",\n \"selectCharacter\": \"请选择一个角色\",\n \"enterChangeReason\": \"请填写形象变化原因\",\n \"enterSubDesc\": \"请填写形象描述\",\n \"insufficientBalance\": \"余额不足\\n\\n{error}\",\n \"designFailed\": \"AI设计失败: {error}\",\n \"addFailed\": \"添加失败: {error}\",\n \"aiDesignPlaceholderNew\": \"例如：一个20岁的女性魔法师，金色长发，蓝色眼睛，穿着紫色法袍...\",\n \"aiDesignPlaceholderSub\": \"例如：换上黑色劲装，脚蹬厚底战靴，准备战斗的状态...\",\n \"aiTipNew\": \"描述你想要的角色，AI会自动生成详细描述\",\n \"aiTipSub\": \"描述角色的新状态，AI会生成子形象描述（只描述变化部分）\",\n \"nameLabel\": \"角色名称\",\n \"namePlaceholder\": \"输入角色名称...\",\n \"descLabel\": \"形象描述\",\n \"descPlaceholder\": \"输入角色形象描述...\",\n \"selectLabel\": \"选择角色\",\n \"selectPlaceholder\": \"-- 请选择角色 --\",\n \"existingAppearances\": \"现有形象：\",\n \"reasonLabel\": \"形象变化原因\",\n \"reasonPlaceholder\": \"例如：换装后、受伤状态、出浴状态...\",\n \"reasonTip\": \"简短描述这个形象与主形象的区别原因\",\n \"subDescPlaceholder\": \"只描述变化部分，例如：换装，脚蹬厚底战靴...\",\n \"subDescTip\": \"子形象只需描述变化部分（服装、状态等），面部和体型会自动继承主形象\",\n \"adding\": \"添加中...\",\n \"insufficientBalanceDefault\": \"账户余额不足，请充值后继续使用\",\n \"addFailedGeneric\": \"添加失败\",\n \"appearancesCount\": \"个形象\",\n \"addCharacter\": \"新增角色\",\n \"addLocation\": \"新增场景\",\n \"aiDesignTip\": \"描述你想要的场景，AI会自动生成名称和详细描述\",\n \"designing\": \"AI设计中...\",\n \"saveName\": \"保存名字\",\n \"saveOnly\": \"仅保存\",\n \"sceneDescription\": \"场景描述\",\n \"scenePrompt\": \"场景描述提示词\",\n \"appearancePrompt\": \"形象描述提示词\",\n \"smartModify\": \"智能修改\",\n \"modifyPlaceholder\": \"例如：改成夜晚，添加月光，增加窗帘...\",\n \"modifyPlaceholderCharacter\": \"例如：把头发改成金色、身高改为180cm、穿黑色西装...\",\n \"modifying\": \"智能修改中...\",\n \"modifyFailed\": \"修改失败\",\n \"editCharacter\": \"编辑角色\",\n \"editLocation\": \"编辑场景\",\n \"saveAndGenerate\": \"保存并生成\",\n \"generatingAutoClose\": \"正在生成图片，完成后将自动关闭...\",\n \"aiLocationTip\": \"输入你想修改的内容，AI会自动调整场景描述\",\n \"aiDesignPlaceholderLocation\": \"例如：一个古老的魔法图书馆，高耸的书架，昏暗的烛光，神秘的氛围...\",\n \"artStyle\": \"画面风格\",\n \"generate\": \"生成\",\n \"introduction\": \"角色介绍\",\n \"introductionPlaceholder\": \"例如：本书主角，第一人称'我'指的就是她。其他角色称她为'小雪'或'雪姐'...\",\n \"introductionTip\": \"描述角色在故事中的身份、叙述视角（如'我'对应谁）、其他角色如何称呼她等\",\n \"saveIntroduction\": \"保存介绍\"\n },\n \"toolbar\": {\n \"filter\": \"筛选\",\n \"viewAll\": \"查看全部\",\n \"showGenerated\": \"已生成\",\n \"showPending\": \"待生成\",\n \"assetManagement\": \"资产管理\",\n \"assetCount\": \"共 {total} 个资产（{appearances} 角色形象 + {locations} 场景）\",\n \"globalAnalyze\": \"全局分析\",\n \"globalAnalyzing\": \"正在执行全局资产分析...\",\n \"globalAnalyzingHint\": \"请勿刷新页面，分析完成后将自动显示结果\",\n \"globalAnalyzingTip\": \"正在分析所有剧集内容，提取角色和场景...\",\n \"globalAnalyzeHint\": \"分析所有剧集内容，提取角色和场景\",\n \"globalAnalyzeSuccess\": \"全局分析完成：新增 {characters} 个角色，{locations} 个场景\",\n \"globalAnalyzeFailed\": \"全局分析失败\",\n \"generateAll\": \"生成全部图片\",\n \"generateAllNoop\": \"所有资产都已有图片，无需生成\",\n \"generating\": \"生成中 ({current}/{total})\",\n \"regenerateAll\": \"重新生成全部\",\n \"regenerateAllConfirm\": \"确定要重新生成所有资产的图片吗？这将覆盖现有图片。\",\n \"noAssetsToGenerate\": \"没有可生成的资产\",\n \"regenerateAllHint\": \"重新生成所有资产图片（覆盖现有）\",\n \"downloadAll\": \"打包下载全部图片\"\n },\n \"common\": {\n \"actions\": \"操作\",\n \"add\": \"添加\",\n \"cancel\": \"取消\",\n \"confirm\": \"确认\",\n \"copy\": \"复制\",\n \"delete\": \"删除\",\n \"download\": \"下载\",\n \"edit\": \"编辑\",\n \"generate\": \"生成\",\n \"generateFailed\": \"生成失败\",\n \"loading\": \"加载中...\",\n \"none\": \"无\",\n \"preview\": \"预览\",\n \"refresh\": \"刷新\",\n \"regenerate\": \"重新生成\",\n \"save\": \"保存\",\n \"status\": \"状态\",\n \"submitFailed\": \"提交失败\",\n \"upload\": \"上传\",\n \"unknownError\": \"未知错误\"\n },\n \"video\": {\n \"panelCard\": {\n \"generating\": \"生成中...\",\n \"editPrompt\": \"编辑提示词\"\n }\n },\n \"smartImport\": {\n \"preview\": {\n \"saving\": \"保存中...\"\n }\n },\n \"storyboard\": {\n \"group\": {\n \"generating\": \"生成中...\"\n }\n },\n \"errors\": {\n \"saveFailed\": \"保存失败，请重试\",\n \"failed\": \"失败，请重试\",\n \"insufficientBalance\": \"账户余额不足\",\n \"aiDesignFailed\": \"AI 设计失败\",\n \"createFailed\": \"创建失败\"\n },\n \"assetLibrary\": {\n \"button\": \"资产库\",\n \"title\": \"资产库\",\n \"copySuccessCharacter\": \"角色形象复制成功\",\n \"copySuccessLocation\": \"场景图片复制成功\",\n \"copySuccessVoice\": \"音色复制成功\",\n \"copyFailed\": \"复制失败: {error}\",\n \"downloadEmpty\": \"当前没有可下载的图片资产\",\n \"downloadFailed\": \"打包下载失败\"\n },\n \"tts\": {\n \"voiceDesignSaved\": \"已为 {name} 设置 AI 设计的声音\",\n \"saveVoiceDesignFailed\": \"保存声音设计失败: {error}\",\n \"title\": \"配音音色\",\n \"noVoice\": \"无音色\",\n \"previewFailed\": \"预览失败: {error}\",\n \"uploadFailed\": \"上传音频失败: {error}\",\n \"uploadQwenHint\": \"上传的音色后续只可使用 IndexTTS 合成，不可用于 QwenTTS。QwenTTS 必须使用 AI 设计音色。\",\n \"uploading\": \"上传中...\",\n \"uploaded\": \"已上传\",\n \"uploadAudio\": \"上传音频\",\n \"pause\": \"暂停\",\n \"preview\": \"试听音色\"\n },\n \"characterProfile\": {\n \"importance\": {\n \"S\": \"S级 - 绝对主角\",\n \"A\": \"A级 - 核心配角\",\n \"B\": \"B级 - 重要配角\",\n \"C\": \"C级 - 次要角色\",\n \"D\": \"D级 - 群众演员\"\n },\n \"costumeLevel\": {\n \"5\": \"皇室/顶奢级\",\n \"4\": \"贵族/精英级\",\n \"3\": \"专业/品质级\",\n \"2\": \"日常/普通级\",\n \"1\": \"朴素/统一级\"\n },\n \"importanceLevel\": \"角色重要性层级\",\n \"characterArchetype\": \"角色原型\",\n \"archetypePlaceholder\": \"如: 霸道总裁、心机婊\",\n \"personalityTags\": \"性格标签\",\n \"addTagPlaceholder\": \"添加标签\",\n \"costumeLevelLabel\": \"服装华丽度\",\n \"suggestedColors\": \"建议色彩\",\n \"colorPlaceholder\": \"如: 深蓝、金色\",\n \"primaryMarker\": \"主要辨识标志\",\n \"markerNote\": \"(S/A级推荐)\",\n \"markingsPlaceholder\": \"如: 眼角泪痣、左耳银色耳钉\",\n \"visualKeywords\": \"视觉关键词\",\n \"keywordsPlaceholder\": \"如: 精英气质、禁欲系\",\n \"editDialogTitle\": \"编辑角色档案 - {name}\",\n \"confirmAndGenerate\": \"确认并生成\",\n \"useExisting\": \"使用已有形象\",\n \"editProfile\": \"编辑档案\",\n \"delete\": \"删除此角色\",\n \"summary\": {\n \"gender\": \"性别:\",\n \"age\": \"年龄:\",\n \"era\": \"时代:\",\n \"class\": \"阶层:\",\n \"occupation\": \"职业:\",\n \"personality\": \"性格:\",\n \"costume\": \"服装:\",\n \"identifier\": \"标志:\"\n },\n \"parseFailed\": \"档案数据解析失败\",\n \"confirmSuccessGenerating\": \"✓ 档案确认成功，正在生成视觉描述\",\n \"confirmFailed\": \"确认失败: {error}\",\n \"noPendingCharacters\": \"没有待确认的角色\",\n \"batchConfirmPrompt\": \"确认为 {count} 个角色生成视觉描述吗？\",\n \"batchConfirmSuccess\": \"✓ 已为 {count} 个角色生成视觉描述\",\n \"batchConfirmFailed\": \"批量确认失败: {error}\",\n \"deleteConfirm\": \"确定要删除此角色吗？此操作不可撤销。\",\n \"deleteSuccess\": \"✓ 角色已删除\",\n \"deleteFailed\": \"删除失败: {error}\"\n },\n \"imageEdit\": {\n \"editCharacterImage\": \"编辑人物图片\",\n \"editLocationImage\": \"编辑场景图片\",\n \"characterLabel\": \"人物: {name}\",\n \"locationLabel\": \"场景: {name}\",\n \"editInstruction\": \"修改指令\",\n \"subtitle\": \"输入修改指令，可选择上传参考图片\",\n \"characterPlaceholder\": \"描述你想要修改的内容，例如：把头发改成金色、添加眼镜、换成休闲装...\",\n \"locationPlaceholder\": \"描述你想要修改的内容，例如：添加更多树木、改成夜晚场景...\",\n \"storyboardPlaceholder\": \"描述你想要修改的内容，例如：改变背景颜色、调整人物表情...\",\n \"noAssetHint\": \"暂无资产，点击\\\"添加资产\\\"选择\",\n \"referenceImages\": \"参考图片\",\n \"referenceImagesHint\": \"(可选，支持粘贴)\",\n \"startEditing\": \"开始编辑\"\n }\n}"
},
{
"path": "messages/zh/auth.json",
"content": "{\n \"welcomeBack\": \"欢迎回来\",\n \"loginTo\": \"登录到waoowaoo\",\n \"createAccount\": \"创建账户\",\n \"joinPlatform\": \"加入waoowaoo\",\n \"phoneNumber\": \"用户名\",\n \"password\": \"密码\",\n \"confirmPassword\": \"确认密码\",\n \"phoneNumberPlaceholder\": \"请输入用户名\",\n \"passwordPlaceholder\": \"请输入密码\",\n \"passwordMinPlaceholder\": \"请输入密码（至少6位）\",\n \"confirmPasswordPlaceholder\": \"请再次输入密码\",\n \"loginButton\": \"登录\",\n \"loginButtonLoading\": \"登录中...\",\n \"signupButton\": \"注册\",\n \"signupButtonLoading\": \"注册中...\",\n \"noAccount\": \"还没有账户？\",\n \"hasAccount\": \"已有账户？\",\n \"signupNow\": \"立即注册\",\n \"signinNow\": \"立即登录\",\n \"backToHome\": \"← 返回首页\",\n \"loginFailed\": \"登录失败，请检查手机号和密码\",\n \"loginError\": \"登录过程中出现错误\",\n \"passwordMismatch\": \"密码确认不匹配\",\n \"passwordTooShort\": \"密码长度至少6位\",\n \"signupSuccess\": \"注册成功！正在跳转到登录页面...\",\n \"signupFailed\": \"注册失败\",\n \"signupError\": \"注册过程中出现错误\",\n \"rateLimited\": \"请求过于频繁，请稍后再试\",\n \"passwordStrength\": {\n \"weak\": \"密码强度：弱 — 建议使用更长的密码，混合大小写、数字和特殊字符\",\n \"fair\": \"密码强度：一般 — 建议增加密码复杂度\",\n \"good\": \"密码强度：良好\",\n \"strong\": \"密码强度：强\"\n }\n}"
},
{
"path": "messages/zh/billing.json",
"content": "{\n \"transactionType\": \"交易类型\",\n \"startDate\": \"开始日期\",\n \"endDate\": \"结束日期\",\n \"all\": \"全部\",\n \"income\": \"收入\",\n \"expense\": \"支出\",\n \"reset\": \"重置\",\n \"filter\": \"筛选\",\n \"noRecords\": \"暂无记录\",\n \"accountRecharge\": \"账户充值\",\n \"serviceConsumption\": \"服务消费\",\n \"balance\": \"余额\",\n \"allTypes\": \"全部类型\"\n}"
},
{
"path": "messages/zh/common.json",
"content": "{\n \"appName\": \"waoowaoo\",\n \"betaVersion\": \"Beta v{version}\",\n \"updateNotice\": {\n \"title\": \"发现新版本\",\n \"subtitle\": \"当前 v{current} · 最新 v{latest}\",\n \"description\": \"软件已发布新版本，建议立刻前去升级。\",\n \"releaseName\": \"版本\",\n \"publishedAt\": \"发布时间\",\n \"updateTag\": \"Update\",\n \"viewRelease\": \"查看更新\",\n \"remindLater\": \"稍后提醒\",\n \"close\": \"关闭\",\n \"openDialog\": \"打开更新弹窗\",\n \"checkUpdate\": \"检查新版本\",\n \"upToDate\": \"已是最新版本\"\n },\n \"loading\": \"加载中...\",\n \"save\": \"保存\",\n \"cancel\": \"取消\",\n \"confirm\": \"确认\",\n \"delete\": \"删除\",\n \"edit\": \"编辑\",\n \"search\": \"搜索\",\n \"clear\": \"清除\",\n \"close\": \"关闭\",\n \"back\": \"返回\",\n \"next\": \"下一步\",\n \"previous\": \"上一步\",\n \"submit\": \"提交\",\n \"reset\": \"重置\",\n \"generate\": \"生成\",\n \"regenerate\": \"重新生成\",\n \"preview\": \"预览\",\n \"download\": \"下载\",\n \"upload\": \"上传\",\n \"select\": \"选择\",\n \"add\": \"添加\",\n \"remove\": \"移除\",\n \"refresh\": \"刷新\",\n \"expand\": \"展开\",\n \"collapse\": \"收起\",\n \"all\": \"全部\",\n \"none\": \"无\",\n \"success\": \"成功\",\n \"error\": \"错误\",\n \"warning\": \"警告\",\n \"info\": \"信息\",\n \"copy\": \"复制\",\n \"paste\": \"粘贴\",\n \"apply\": \"应用\",\n \"autoSave\": \"自动保存\",\n \"saved\": \"已保存\",\n \"episode\": \"剧集\",\n \"project\": \"项目\",\n \"editEpisodeName\": \"编辑剧集名称\",\n \"deleteEpisode\": \"删除剧集\",\n \"deleteEpisodeConfirm\": \"确认删除\",\n \"newEpisode\": \"新建剧集\",\n \"optional\": \"（可选）\",\n \"rename\": \"重命名\",\n \"dragToReorder\": \"拖动调整位置\",\n \"episodeNamePlaceholder\": \"输入剧集名称...\",\n \"cancelSelection\": \"取消选择\",\n \"referenceImage\": \"参考图\",\n \"previewLarge\": \"预览大图\",\n \"viewOriginal\": \"查看原图\",\n \"schemeN\": \"方案 {n}\",\n \"insufficientBalance\": \"余额不足\",\n \"insufficientBalanceDetail\": \"账户余额不足，请充值后继续使用\",\n \"operationFailed\": \"操作失败\",\n \"pleaseRetry\": \"请重试\",\n \"recommended\": \"推荐\",\n \"language\": {\n \"select\": \"选择语言\",\n \"zh\": \"中文\",\n \"en\": \"English\",\n \"switchConfirmTitle\": \"切换语言？\",\n \"switchConfirmMessage\": \"切换到 {targetLanguage} 后，不仅界面文字会改变，整条流程的提示词模板、剧本生成和任务输出语言也会同步切换。是否继续？\",\n \"switchConfirmAction\": \"确认切换\"\n },\n \"taskStatus\": {\n \"intent\": {\n \"generate\": {\n \"running\": {\n \"image\": \"生成中\",\n \"video\": \"生成中\",\n \"audio\": \"生成中\",\n \"text\": \"生成中\"\n }\n },\n \"regenerate\": {\n \"running\": {\n \"image\": \"重新生成中\",\n \"video\": \"重新生成中\",\n \"audio\": \"重新生成中\",\n \"text\": \"重新生成中\"\n }\n },\n \"modify\": {\n \"running\": {\n \"image\": \"修改中\",\n \"video\": \"修改中\",\n \"audio\": \"修改中\",\n \"text\": \"修改中\"\n }\n },\n \"analyze\": {\n \"running\": {\n \"image\": \"分析中\",\n \"video\": \"分析中\",\n \"audio\": \"分析中\",\n \"text\": \"分析中\"\n }\n },\n \"build\": {\n \"running\": {\n \"image\": \"构建中\",\n \"video\": \"构建中\",\n \"audio\": \"构建中\",\n \"text\": \"构建中\"\n }\n },\n \"convert\": {\n \"running\": {\n \"image\": \"转换中\",\n \"video\": \"转换中\",\n \"audio\": \"转换中\",\n \"text\": \"转换中\"\n }\n },\n \"process\": {\n \"running\": {\n \"image\": \"处理中\",\n \"video\": \"处理中\",\n \"audio\": \"处理中\",\n \"text\": \"处理中\"\n }\n }\n },\n \"failed\": {\n \"image\": \"处理失败\",\n \"video\": \"处理失败\",\n \"audio\": \"处理失败\",\n \"text\": \"处理失败\"\n }\n }\n}"
},
{
"path": "messages/zh/configModal.json",
"content": "{\n \"title\": \"项目配置\",\n \"subtitle\": \"默认沿用设置中心的全局配置，也可为当前项目单独自定义，修改仅对本项目生效。\",\n \"saved\": \"已保存\",\n \"autoSave\": \"自动保存\",\n \"visualStyle\": \"视觉风格\",\n \"modelParams\": \"模型参数\",\n \"aspectRatio\": \"画面比例\",\n \"ttsSettings\": \"旁白配置\",\n \"loadingModels\": \"加载模型列表...\",\n \"analysisModel\": \"分析模型\",\n \"characterModel\": \"人物生成模型\",\n \"locationModel\": \"场景生成模型\",\n \"storyboardModel\": \"分镜图像模型\",\n \"editModel\": \"修图/编辑模型\",\n \"videoModel\": \"视频模型\",\n \"audioModel\": \"语音合成模型\",\n \"videoResolution\": \"视频分辨率\",\n \"ttsVoice\": \"旁白音色\",\n \"ttsRate\": \"语速\",\n \"fetchModelsFailed\": \"获取用户模型列表失败\",\n \"placeholder\": \"请输入...\",\n \"description\": \"描述\",\n \"hint\": \"提示\",\n \"pleaseSelect\": \"请选择...\",\n \"selectModel\": \"选择模型\",\n \"paramConfig\": \"参数配置\",\n \"fixed\": \"固定\",\n \"noParams\": \"该模型无可配置参数\",\n \"confirm\": \"确认\",\n \"cancel\": \"取消\",\n \"delete\": \"删除\",\n \"boolOn\": \"开\",\n \"boolOff\": \"关\"\n}\n"
},
{
"path": "messages/zh/errors.json",
"content": "{\n \"UNAUTHORIZED\": \"请先登录\",\n \"FORBIDDEN\": \"没有权限访问\",\n \"NOT_FOUND\": \"资源不存在\",\n \"INSUFFICIENT_BALANCE\": \"API余额不足，请充值后重试\",\n \"RATE_LIMIT\": \"请求过于频繁，请 {retryAfter} 秒后重试\",\n \"MODEL_NOT_OPEN\": \"模型权限未开通。请前往 https://console.volcengine.com/ark/region:ark+cn-beijing/openManagement?LLM=%7B%7D&advancedActiveKey=model ，在模型管理页面点击右上角「一键开通所有模型」\",\n \"MODEL_NOT_REGISTERED\": \"模型未注册，请先在配置中添加可用模型\",\n \"MODEL_NOT_CONFIGURED\": \"未配置可用模型，请先前往「设置」页面添加对应类型的模型后再试\",\n \"QUOTA_EXCEEDED\": \"配额已用尽，请稍后重试\",\n \"GENERATION_FAILED\": \"生成失败，请重试\",\n \"GENERATION_TIMEOUT\": \"生成超时，请重试\",\n \"SENSITIVE_CONTENT\": \"内容可能包含敏感信息\",\n \"INVALID_PARAMS\": \"参数错误\",\n \"MISSING_CONFIG\": \"请先完成模型配置\",\n \"INTERNAL_ERROR\": \"服务器错误，请稍后重试\",\n \"NETWORK_ERROR\": \"网络错误，请检查连接\",\n \"EMPTY_RESPONSE\": \"模型返回空响应（无有效内容），请稍后重试\",\n \"EXTERNAL_ERROR\": \"外部服务暂时不可用，请稍后重试\",\n \"TASK_NOT_READY\": \"任务正在处理中\",\n \"NO_RESULT\": \"任务无结果\",\n \"CONFLICT\": \"资源状态冲突\"\n}"
},
{
"path": "messages/zh/landing.json",
"content": "{\n \"title\": \"waoowaoo\",\n \"subtitle\": \"AI影视Studio\",\n \"enterWorkspace\": \"进入工作区\",\n \"loading\": \"加载中...\",\n \"getStarted\": \"立即体验\",\n \"learnMore\": \"了解更多\",\n \"features\": {\n \"title\": \"释放无限创造力\",\n \"subtitle\": \"全流程 AI 辅助，从剧本到成片\",\n \"character\": {\n \"title\": \"角色工坊\",\n \"description\": \"打造独一无二的动漫角色，保持高度一致性\"\n },\n \"storyboard\": {\n \"title\": \"智能分镜\",\n \"description\": \"文字一键转分镜，精准把控叙事节奏\"\n },\n \"world\": {\n \"title\": \"世界构建\",\n \"description\": \"沉浸式场景生成，构建宏大的故事背景\"\n }\n },\n \"footer\": {\n \"copyright\": \"2026 waoowaoo AI. All rights reserved.\"\n }\n}"
},
{
"path": "messages/zh/layout.json",
"content": "{\n \"title\": \"AI动漫制作工作台\",\n \"description\": \"使用最先进的AI技术创建专业级动漫内容\"\n}\n"
},
{
"path": "messages/zh/modelSection.json",
"content": "{\n \"llmModels\": \"文本模型列表\",\n \"imageModels\": \"图片模型列表\",\n \"videoModels\": \"视频模型列表\",\n \"price\": \"价格\",\n \"pricePerMillion\": \"每百万token\",\n \"pricePerImage\": \"每张\",\n \"pricePerVideo\": \"每条\",\n \"name\": \"名称\",\n \"modelId\": \"模型 ID\",\n \"modelName\": \"模型名称\",\n \"provider\": \"厂商\",\n \"resolution\": \"分辨率\",\n \"add\": \"添加\",\n \"addModel\": \"添加模型\",\n \"addNewModel\": \"添加新模型\",\n \"selectPreset\": \"选择预设模型\",\n \"customModel\": \"自定义模型\",\n \"confirmAdd\": \"确认添加\",\n \"cancel\": \"取消\",\n \"done\": \"完成\",\n \"fillComplete\": \"请填写完整信息\",\n \"noModels\": \"暂无模型，点击上方按钮添加\",\n \"noApiKey\": \"请配置API Key\",\n \"batchMode\": \"批量\",\n \"batchModeTooltip\": \"离线推理，价格便宜50%，24小时内完成\"\n}\n"
},
{
"path": "messages/zh/nav.json",
"content": "{\n \"workspace\": \"工作区\",\n \"assetHub\": \"资产中心\",\n \"profile\": \"设置中心\",\n \"downloadLogs\": \"下载日志\",\n \"signin\": \"登录\",\n \"signup\": \"注册\",\n \"logout\": \"退出登录\"\n}\n"
},
{
"path": "messages/zh/novel-promotion.json",
"content": "{\n \"stages\": {\n \"story\": \"故事\",\n \"script\": \"剧本\",\n \"storyboard\": \"分镜\",\n \"video\": \"成片\",\n \"editor\": \"AI剪辑\",\n \"editorComingSoon\": \"开发中，关注我们获取最新消息\"\n },\n \"buttons\": {\n \"assetLibrary\": \"资产库\",\n \"settings\": \"项目配置\",\n \"refreshData\": \"刷新项目数据\",\n \"enterVideoGeneration\": \"进入视频生成 →\"\n },\n \"smartImport\": {\n \"title\": \"开启你的创作之旅\",\n \"subtitle\": \"首先，选择你的创作方式\",\n \"manualCreate\": {\n \"title\": \"从第一集开始创作\",\n \"description\": \"从第一集开始，适合边写边播或单集短视频制作\",\n \"button\": \"开始创作\"\n },\n \"smartImport\": {\n \"title\": \"智能全书导入\",\n \"description\": \"上传整本小说或剧本，AI 引擎自动识别章节结构，一键完成智能分集。\",\n \"button\": \"立即导入\",\n \"recommended\": \"推荐\"\n },\n \"upload\": {\n \"title\": \"上传原始素材\",\n \"subtitle\": \"AI 引擎已准备就绪，一键自动分集与格式化\",\n \"maxWords\": \"（最大支持 3 万字）\",\n \"textInput\": \"输入文本内容\",\n \"documentUpload\": \"上传完整文档\",\n \"placeholder\": \"在此处粘贴你的小说章节或剧本内容...\",\n \"filePlaceholder\": \"已上传文件模式\",\n \"clickUpload\": \"点击上传文档\",\n \"clearTextFirst\": \"请先清空左侧文本\",\n \"supportedFormats\": \"支持 Word, TXT 格式\",\n \"preview\": \"预览\",\n \"expandPreview\": \"展开更多\",\n \"collapsePreview\": \"收起预览\",\n \"deleteFile\": \"删除文件\",\n \"startAnalysis\": \"开始智能分析\",\n \"back\": \"返回\",\n \"words\": \"字\"\n },\n \"analyzing\": {\n \"title\": \"AI 正在分析你的故事\",\n \"description\": \"识别章节结构，智能分集中...\",\n \"autoSave\": \"分析完成后将自动保存\"\n },\n \"preview\": {\n \"title\": \"智能分集完成\",\n \"episodeCount\": \"已为你自动分为 {count} 集\",\n \"totalWords\": \"总计 {count} 字\",\n \"autoSaved\": \"✓ 已自动保存\",\n \"reanalyze\": \"重新分析\",\n \"confirm\": \"确认完成\",\n \"saving\": \"保存中...\",\n \"episodeList\": \"剧集列表\",\n \"addEpisode\": \"添加剧集\",\n \"averageWords\": \"平均每集\",\n \"episodeContent\": \"剧集内容\",\n \"episodePlaceholder\": \"输入剧集标题...\",\n \"summaryPlaceholder\": \"输入剧情简介...\",\n \"newEpisode\": \"新剧集\",\n \"deleteEpisode\": \"删除剧集\",\n \"deleteConfirm\": {\n \"title\": \"确认删除\",\n \"message\": \"确定要删除「{title}」吗？\",\n \"cancel\": \"取消\",\n \"confirm\": \"确认删除\"\n },\n \"tip\": {\n \"title\": \"提示\",\n \"content\": \"你可以直接编辑标题、简介和内容。点击【确认完成】后，剧集将正式导入到项目中\"\n }\n },\n \"errors\": {\n \"fileTooLarge\": \"文件过大，请上传小于 10MB 的文件\",\n \"docNotSupported\": \"不支持 .doc 格式，请使用 Word 转换为 .docx\",\n \"fileEmpty\": \"文件内容为空\",\n \"fileReadError\": \"文件读取失败，请重试\",\n \"uploadFirst\": \"请先上传或粘贴内容\",\n \"analyzeFailed\": \"分析失败\",\n \"saveFailed\": \"保存失败\"\n },\n \"cancelConfirm\": \"确定要取消吗？已分析的剧集将被清空。\"\n },\n \"storyInput\": {\n \"currentEditing\": \"当前正在编辑：{name}\",\n \"editingTip\": \"以下制作流程仅针对本集,如有其他剧集请在左上角切换\",\n \"wordCount\": \"字数：\",\n \"assetLibraryTip\": {\n \"title\": \"需要自定义角色和场景？\",\n \"description\": \"点击右上角的「资产库」按钮，可以上传资产设定文档或手动添加角色/场景。AI 将优先使用资产库中的设定进行分析。\"\n },\n \"videoRatio\": \"画面比例\",\n \"videoRatioHint\": \"选择合适的画面比例，可以更好适配投放平台和素材形态\",\n \"ratioUsage\": {\n \"1_1\": \"1:1：方形画面，适合头像、方图封面、社交平台通用封面\",\n \"9_16\": \"9:16：竖屏视频，适合抖音、快手、视频号等短视频平台\",\n \"16_9\": \"16:9：横屏视频，适合 B 站、YouTube 等长视频平台，以及电脑端播放\",\n \"4_3\": \"4:3：传统电视比例，适合部分老素材或需要保守裁切的内容\",\n \"3_4\": \"3:4：略偏竖的画面，适合图文 + 视频混排场景\",\n \"2_3\": \"2:3：偏竖直画面，适合海报、人物立绘等内容\",\n \"3_2\": \"3:2：略偏横的画面，适合风景、剧情类视频\",\n \"4_5\": \"4:5：竖幅海报比例，适合社交平台信息流图片\",\n \"5_4\": \"5:4：横幅海报比例，适合 PC 端 Banner 等\",\n \"21_9\": \"21:9：超宽影院级画幅，适合电影感大片和全景镜头\"\n },\n \"ratioUsageTag\": {\n \"1_1\": \"方形 · 头像/封面\",\n \"9_16\": \"竖屏 · 短视频\",\n \"16_9\": \"横屏 · 长视频\",\n \"4_3\": \"横屏 · 传统电视\",\n \"3_4\": \"竖屏 · 图文混排\",\n \"2_3\": \"竖屏 · 海报/立绘\",\n \"3_2\": \"横屏 · 风景/剧情\",\n \"4_5\": \"竖屏 · 信息流图\",\n \"5_4\": \"横屏 · Banner\",\n \"21_9\": \"超宽 · 电影感\"\n },\n \"visualStyle\": \"视觉风格\",\n \"visualStyleHint\": \"根据受众选择画面风格，例如：真人风格适合写实剧情，动漫风格适合二次元内容\",\n \"currentConfigSummary\": \"当前配置：{ratio} · {style}，后续生成都会使用此组合\",\n \"assetLibraryRatioNote\": \"资产库比例不受影响\",\n \"moreConfig\": \"更多配置请点击右上角「 配置」按钮\",\n \"narration\": {\n \"title\": \"启用旁白配音\",\n \"description\": \"生成 TTS 语音旁白，为视频添加解说\"\n },\n \"creating\": \"AI 创作中...\",\n \"ready\": \"✓ 配置完成，可以进入下一步\",\n \"pleaseInput\": \"请先输入剧本内容\"\n },\n \"execution\": {\n \"selectEpisode\": \"请先选择剧集\",\n \"fillContentFirst\": \"请先填写内容\",\n \"requestAborted\": \"请求已中断（可能因页面刷新）\",\n \"analysisFailed\": \"资产分析失败\",\n \"prepareFailed\": \"准备失败\",\n \"generationFailed\": \"生成失败\",\n \"batchVideoFailed\": \"批量生成视频失败\",\n \"updateFailed\": \"更新失败\",\n \"saveFailed\": \"保存失败\",\n \"storyToScriptRunning\": \"Story→Script V2 运行中\",\n \"scriptToStoryboardRunning\": \"Script→Storyboard V2 运行中\",\n \"storyToScriptFailed\": \"内容转剧本失败\",\n \"scriptToStoryboardFailed\": \"剧本转分镜失败\",\n \"taskStreamTimeout\": \"任务执行超时，请检查任务是否仍在运行，或重新触发\"\n },\n \"rebuildConfirm\": {\n \"storyToScript\": {\n \"title\": \"将重建剧本流程\",\n \"message\": \"检测到当前剧集已有下游分镜数据（{storyboardCount} 个分镜，{panelCount} 个镜头面板）。继续执行将清空这些数据并重新生成，是否继续？\"\n },\n \"scriptToStoryboard\": {\n \"title\": \"将重建分镜数据\",\n \"message\": \"检测到当前剧集已有分镜数据（{storyboardCount} 个分镜，{panelCount} 个镜头面板）。继续执行将清空当前分镜并重新生成，是否继续？\"\n },\n \"confirm\": \"继续并清空\",\n \"cancel\": \"取消\"\n }\n}"
},
{
"path": "messages/zh/profile.json",
"content": "{\n \"user\": \"用户\",\n \"personalAccount\": \"个人账户\",\n \"availableBalance\": \"可用余额\",\n \"openSourceNoBilling\": \"开源版本，无需计费\",\n \"frozen\": \"冻结\",\n \"totalSpent\": \"已消费\",\n \"apiConfig\": \"API 配置\",\n \"rechargeRecords\": \"充值记录\",\n \"billingRecords\": \"扣费记录\",\n \"logout\": \"退出登录\",\n \"downloadLogs\": \"下载日志\",\n \"accountTransactions\": \"账户流水\",\n \"projectDetails\": \"项目明细\",\n \"summary\": \"汇总\",\n \"transactions\": \"流水\",\n \"noTransactions\": \"暂无流水记录\",\n \"noProjectCosts\": \"暂无项目费用记录\",\n \"noDetails\": \"该项目暂无费用明细\",\n \"noRecords\": \"暂无记录\",\n \"byType\": \"按类型\",\n \"byAction\": \"按操作\",\n \"times\": \"次\",\n \"total\": \"总计\",\n \"filter\": \"筛选\",\n \"allTypes\": \"全部类型\",\n \"recharge\": \"账户充值\",\n \"consume\": \"服务消费\",\n \"balanceAfter\": \"余额 {amount}\",\n \"recordCount\": \"{count} 条记录\",\n \"totalCost\": \"总计 {amount}\",\n \"previousPage\": \"上一页\",\n \"nextPage\": \"下一页\",\n \"pagination\": \"共 {total} 条，第 {page} / {totalPages} 页\",\n \"episodeLabel\": \"第 {number} 集\",\n \"billingDetail\": {\n \"imageWithRes\": \"{count}张 · {resolution}\",\n \"image\": \"{count}张\",\n \"videoWithRes\": \"{count}段 · {resolution}\",\n \"video\": \"{count}段\",\n \"tokens\": \"{count} tokens\",\n \"seconds\": \"{count}秒\",\n \"calls\": \"{count}次\"\n },\n \"apiTypes\": {\n \"image\": \"图片生成\",\n \"video\": \"视频生成\",\n \"text\": \"文本分析\",\n \"tts\": \"语音合成\",\n \"voice\": \"配音\",\n \"voice_design\": \"声音设计\",\n \"lip_sync\": \"口型同步\"\n },\n \"actionTypes\": {\n \"image_panel\": \"分镜生图\",\n \"image_character\": \"角色生图\",\n \"image_location\": \"场景生图\",\n \"video_panel\": \"视频生成\",\n \"lip_sync\": \"口型同步\",\n \"voice_line\": \"配音合成\",\n \"voice_design\": \"声音设计\",\n \"asset_hub_voice_design\": \"素材库声音设计\",\n \"regenerate_storyboard_text\": \"重生成分镜文案\",\n \"insert_panel\": \"插入面板\",\n \"panel_variant\": \"镜头变体\",\n \"modify_asset_image\": \"修图\",\n \"regenerate_group\": \"批量重生成\",\n \"asset_hub_image\": \"素材库生图\",\n \"asset_hub_modify\": \"素材库修图\",\n \"analyze_novel\": \"小说分析\",\n \"story_to_script_run\": \"故事转剧本\",\n \"script_to_storyboard_run\": \"剧本转分镜\",\n \"clips_build\": \"片段合成\",\n \"screenplay_convert\": \"剧本转换\",\n \"voice_analyze\": \"声音分析\",\n \"analyze_global\": \"全局分析\",\n \"ai_modify_appearance\": \"AI 修改角色形象\",\n \"ai_modify_location\": \"AI 修改场景\",\n \"ai_modify_shot_prompt\": \"AI 修改镜头提示词\",\n \"analyze_shot_variants\": \"镜头变体分析\",\n \"ai_create_character\": \"AI 创建角色\",\n \"ai_create_location\": \"AI 创建场景\",\n \"reference_to_character\": \"参考图转角色\",\n \"character_profile_confirm\": \"确认角色档案\",\n \"character_profile_batch_confirm\": \"批量确认角色档案\",\n \"episode_split_llm\": \"章节拆分\",\n \"asset_hub_ai_design_character\": \"素材库 AI 设计角色\",\n \"asset_hub_ai_design_location\": \"素材库 AI 设计场景\",\n \"asset_hub_ai_modify_character\": \"素材库 AI 修改角色\",\n \"asset_hub_ai_modify_location\": \"素材库 AI 修改场景\",\n \"asset_hub_reference_to_character\": \"素材库参考图转角色\",\n \"storyboard\": \"分镜图\",\n \"storyboard_candidate\": \"分镜候选\",\n \"character\": \"角色图\",\n \"location\": \"场景图\",\n \"video\": \"视频\",\n \"analyze\": \"分析\",\n \"analyze_character\": \"角色分析\",\n \"analyze_location\": \"场景分析\",\n \"clips\": \"片段切割\",\n \"storyboard_text_plan\": \"分镜规划\",\n \"storyboard_text_detail\": \"分镜细节\",\n \"tts\": \"语音合成\",\n \"regenerate\": \"重生成\",\n \"voice-generate\": \"配音生成\",\n \"voice-design\": \"声音设计\",\n \"lip-sync\": \"口型同步\"\n }\n}"
},
{
"path": "messages/zh/progress.json",
"content": "{\n \"analyzing\": \"正在分析故事结构...\",\n \"splittingClips\": \"正在分割片段...\",\n \"convertingScreenplay\": \"正在转换为分镜脚本...\",\n \"submittingStoryboard\": \"正在提交分镜...\",\n \"step\": \"第 {current} 步，共 {total} 步\",\n \"status\": {\n \"completed\": \"已完成\",\n \"failed\": \"失败\",\n \"processing\": \"进行中\",\n \"queued\": \"排队中\",\n \"pending\": \"未开始\"\n },\n \"stageCard\": {\n \"stage\": \"阶段\",\n \"realtimeStream\": \"实时流\",\n \"currentStage\": \"当前阶段\",\n \"outputTitle\": \"AI 实时输出 · {stage}\",\n \"waitingModelOutput\": \"等待模型输出...\",\n \"reasoningNotProvided\": \"该步骤未返回思考过程\"\n },\n \"runtime\": {\n \"waitingExecution\": \"等待执行\",\n \"taskCreated\": \"任务已创建\",\n \"taskStarted\": \"任务开始处理\",\n \"taskCompleted\": \"任务已完成\",\n \"taskFailed\": \"任务失败\",\n \"taskProcessing\": \"任务处理中...\",\n \"llm\": {\n \"processing\": \"模型处理中...\",\n \"output\": \"模型正在输出...\",\n \"reasoning\": \"模型正在推理...\",\n \"completed\": \"模型输出完成\",\n \"failed\": \"模型输出失败\"\n },\n \"stage\": {\n \"llmSubmit\": \"模型请求提交中\",\n \"llmStreaming\": \"模型流式输出中\",\n \"llmFallbackNonStream\": \"模型降级为非流式模式\",\n \"llmCompleted\": \"模型输出完成\",\n \"llmFailed\": \"模型输出失败\"\n }\n },\n \"taskType\": {\n \"generic\": \"任务\",\n \"imagePanel\": \"分镜图片\",\n \"imageCharacter\": \"角色图片\",\n \"imageLocation\": \"场景图片\",\n \"videoPanel\": \"视频生成\",\n \"lipSync\": \"口型同步\",\n \"voiceLine\": \"配音生成\",\n \"voiceDesign\": \"声音设计\",\n \"assetHubVoiceDesign\": \"资产库声音设计\",\n \"regenerateStoryboardText\": \"重生成分镜文本\",\n \"insertPanel\": \"插入分镜\",\n \"panelVariant\": \"分镜变体\",\n \"modifyAssetImage\": \"图片编辑\",\n \"regenerateGroup\": \"批量重生成\",\n \"assetHubImage\": \"资产中心图片\",\n \"assetHubModify\": \"资产中心编辑\",\n \"analyzeNovel\": \"内容分析\",\n \"storyToScriptRun\": \"剧本拆解\",\n \"scriptToStoryboardRun\": \"分镜生成\",\n \"clipsBuild\": \"片段生成\",\n \"screenplayConvert\": \"剧本转换\",\n \"voiceAnalyze\": \"台词分析\",\n \"analyzeGlobal\": \"全局分析\",\n \"aiModifyAppearance\": \"角色描述修改\",\n \"aiModifyLocation\": \"场景描述修改\",\n \"aiModifyShotPrompt\": \"镜头提示词修改\",\n \"analyzeShotVariants\": \"镜头变体分析\",\n \"aiCreateCharacter\": \"项目角色设计\",\n \"aiCreateLocation\": \"项目场景设计\",\n \"referenceToCharacter\": \"参考图转角色\",\n \"characterProfileConfirm\": \"角色档案确认\",\n \"characterProfileBatchConfirm\": \"批量角色档案确认\",\n \"episodeSplitLlm\": \"智能分集\",\n \"assetHubAiDesignCharacter\": \"资产库角色设计\",\n \"assetHubAiDesignLocation\": \"资产库场景设计\",\n \"assetHubAiModifyCharacter\": \"资产库角色修改\",\n \"assetHubAiModifyLocation\": \"资产库场景修改\",\n \"assetHubReferenceToCharacter\": \"资产库参考图转角色\"\n },\n \"stage\": {\n \"received\": \"任务已接收\",\n \"generateCharacterImage\": \"生成角色图片\",\n \"generateLocationImage\": \"生成场景图片\",\n \"generatePanelCandidate\": \"生成候选分镜图\",\n \"generatePanelVideo\": \"生成分镜视频\",\n \"generateVoiceSubmit\": \"提交配音任务\",\n \"generateVoicePersist\": \"保存配音结果\",\n \"voiceDesignSubmit\": \"提交声音设计任务\",\n \"voiceDesignDone\": \"声音设计完成\",\n \"submitLipSync\": \"提交口型同步任务\",\n \"persistLipSync\": \"保存口型同步结果\",\n \"storyboardClip\": \"生成片段分镜\",\n \"regenerateStoryboardPrepare\": \"准备重生成分镜\",\n \"regenerateStoryboardPersist\": \"保存重生成分镜\",\n \"storyToScriptPrepare\": \"准备剧本拆解参数\",\n \"storyToScriptStep\": \"执行剧本拆解步骤\",\n \"storyToScriptPersist\": \"保存剧本拆解结果\",\n \"storyToScriptPersistDone\": \"剧本拆解结果已保存\",\n \"scriptToStoryboardPrepare\": \"准备分镜生成参数\",\n \"scriptToStoryboardStep\": \"执行分镜生成步骤\",\n \"scriptToStoryboardPersist\": \"保存分镜结果\",\n \"scriptToStoryboardPersistDone\": \"分镜与台词结果已保存\",\n \"insertPanelGenerateText\": \"生成插入镜头文本\",\n \"insertPanelPersist\": \"保存插入镜头\",\n \"pollingExternal\": \"等待外部服务返回\",\n \"enqueueFailed\": \"任务入队失败\",\n \"llmProxySubmit\": \"提交 LLM 任务\",\n \"llmProxyExecute\": \"执行 LLM 任务\",\n \"llmProxyPersist\": \"保存 LLM 结果\"\n },\n \"runConsole\": {\n \"storyToScript\": \"内容到剧本\",\n \"scriptToStoryboard\": \"剧本到分镜\",\n \"storyToScriptRunning\": \"Story→Script 运行中\",\n \"scriptToStoryboardRunning\": \"Script→Storyboard 运行中\",\n \"storyToScriptSubtitle\": \"Story To Script V2\",\n \"scriptToStoryboardSubtitle\": \"Script To Storyboard V2\",\n \"stop\": \"停止\",\n \"minimize\": \"最小化\"\n },\n \"streamStep\": {\n \"analyzeCharacters\": \"角色分析\",\n \"analyzeLocations\": \"场景分析\",\n \"splitClips\": \"片段切分\",\n \"screenplayConversion\": \"剧本转换\",\n \"storyboardPlan\": \"分镜规划\",\n \"cinematographyRules\": \"摄影规则生成\",\n \"actingDirection\": \"演技指导生成\",\n \"storyboardDetailRefine\": \"分镜细节补全\",\n \"voiceAnalyze\": \"台词分析\"\n }\n}\n"
},
{
"path": "messages/zh/providerSection.json",
"content": "{\n \"addProvider\": \"+ 添加提供商\",\n \"name\": \"名称\",\n \"add\": \"添加\",\n \"save\": \"保存\",\n \"fillRequired\": \"请填写必要信息\"\n}"
},
{
"path": "messages/zh/scriptView.json",
"content": "{\n \"title\": \"剧本视图\",\n \"scriptBreakdown\": \"剧本拆解\",\n \"splitCount\": \"已拆分 {count} 个分镜\",\n \"noClips\": \"暂无分镜，请先在故事视图生成\",\n \"segment\": {\n \"title\": \"片段 {index}\",\n \"selected\": \"(选中)\"\n },\n \"inSceneAssets\": \"剧中资产\",\n \"currentSelected\": \"当前选中: 片段 {number}\",\n \"assetView\": {\n \"allClips\": \"全部片段\",\n \"viewingClip\": \"查看片段 {number}\"\n },\n \"asset\": {\n \"generateCharacter\": \"点击生成形象 →\",\n \"generateLocation\": \"点击生成场景 →\",\n \"removeCharacterConfirm\": \"确定要从当前片段移除该角色吗？\",\n \"removeLocationConfirm\": \"确定要从当前片段移除该场景吗？\",\n \"removeFromClip\": \"从当前片段移除\",\n \"noAudio\": \"无音频\",\n \"playing\": \"播放中\",\n \"listen\": \"试听\",\n \"activeCharacters\": \"出场角色\",\n \"activeLocations\": \"出场场景\",\n \"selectCharacter\": \"选择要添加的角色/形象\",\n \"selectLocation\": \"选择要添加的场景\",\n \"loadingAssets\": \"加载资产中...\",\n \"appearanceCount\": \"{count} 个形象\",\n \"added\": \"已添加\",\n \"primary\": \"主形象\",\n \"subAppearance\": \"子形象\",\n \"defaultAppearance\": \"初始形象\",\n \"clickToRemove\": \"点击移除 {name}\",\n \"clickToAdd\": \"点击添加 {name}\"\n },\n \"screenplay\": {\n \"scene\": \"场景 {number}\",\n \"location\": \"场景：\",\n \"locationTime\": \"时间：\",\n \"day\": \"日\",\n \"night\": \"夜\",\n \"dawn\": \"晨\",\n \"dusk\": \"昏\",\n \"dialogue\": \"对白\",\n \"action\": \"动作\",\n \"narration\": \"旁白\",\n \"content\": \"原文内容\",\n \"noContent\": \"暂无内容\",\n \"clickToEdit\": \"点击编辑\",\n \"interior\": \"内景\",\n \"exterior\": \"外景\",\n \"characters\": \"出场角色\",\n \"noCharacter\": \"暂无角色信息\",\n \"noLocation\": \"暂无出场场景\",\n \"noCharacterInClip\": \"暂无出场角色\"\n },\n \"confirm\": {\n \"removeCharacter\": \"确定要从当前片段移除该角色吗？\",\n \"removeLocation\": \"确定要从当前片段移除该场景吗？\"\n },\n \"generate\": {\n \"missingAssets\": \"还有 {count} 个资产未生成形象\",\n \"missingAssetsTip\": \"请先在\",\n \"missingAssetsTipLink\": \"中生成所有出场角色和场景的形象\",\n \"generating\": \"正在生成画面...\",\n \"startGenerate\": \"确认并开始绘制 →\"\n }\n}"
},
{
"path": "messages/zh/smartImport.json",
"content": "{\n \"title\": \"开启你的创作之旅\",\n \"subtitle\": \"首先，选择你的创作方式\",\n \"manualCreate\": {\n \"title\": \"从第一集开始创作\",\n \"description\": \"从第一集开始，适合边写边播或单集短视频制作\",\n \"button\": \"开始创作\"\n },\n \"manualDesc\": \"从第一集开始，适合边写边播或单集短视频制作\",\n \"startCreate\": \"开始创作\",\n \"smartImport\": {\n \"title\": \"智能文本分集\",\n \"description\": \"上传整本小说或剧本，AI 引擎自动识别章节结构，一键完成智能分集。\",\n \"button\": \"立即导入\",\n \"recommended\": \"推荐\"\n },\n \"markerDetection\": {\n \"enable\": \"优先使用章节标记（如第X集/Episode X）\",\n \"tooltip\": \"自动识别【第X集/章】【Episode/Chapter X】等标记，免费快速\"\n },\n \"smartImportDesc\": \"上传整本小说或剧本，AI 引擎自动识别章节结构，一键完成智能分集。\",\n \"recommended\": \"推荐\",\n \"importNow\": \"立即导入\",\n \"uploadTitle\": \"上传原始素材\",\n \"uploadSubtitle\": \"AI 引擎已准备就绪，一键自动分集与格式化\",\n \"maxWords\": \"最大支持 3 万字\",\n \"textInput\": \"输入文本内容\",\n \"textPlaceholder\": \"在此处粘贴你的小说章节或剧本内容...\",\n \"uploadDoc\": \"上传完整文档\",\n \"clickUpload\": \"点击上传文档\",\n \"clearText\": \"请先清空左侧文本\",\n \"supportFormat\": \"支持 Word, TXT 格式\",\n \"fileMax\": \"文件最大 3万字\",\n \"words\": \"字\",\n \"startAnalyzing\": \"开始分析\",\n \"analyzing\": {\n \"title\": \"AI 正在分析你的故事\",\n \"description\": \"识别章节结构，智能分集中...\",\n \"autoSave\": \"分析完成后将自动保存\"\n },\n \"analyzingDesc\": \"识别章节结构，智能分集中...\",\n \"autoSave\": \"分析完成后将自动保存\",\n \"splitComplete\": \"智能分集完成\",\n \"splitResult\": \"已为你自动分为 {count} 集，总计 {words} 字\",\n \"saved\": \"已自动保存\",\n \"reAnalyze\": \"重新分析\",\n \"confirmComplete\": \"确认完成\",\n \"saving\": \"保存中...\",\n \"episodeList\": \"剧集列表\",\n \"episodes\": \"集\",\n \"episode\": \"第 {num} 集\",\n \"addEpisode\": \"添加剧集\",\n \"newEpisode\": \"新剧集\",\n \"avgWords\": \"平均每集\",\n \"episodeContent\": \"剧集内容\",\n \"plotSummary\": \"剧情简介\",\n \"enterTitle\": \"输入剧集标题...\",\n \"enterSummary\": \"输入剧情简介...\",\n \"confirmDelete\": \"确认删除\",\n \"deleteConfirmMsg\": \"确定要删除「{title}」吗？\",\n \"preview\": {\n \"title\": \"智能分集完成\",\n \"episodeCount\": \"已为你自动分为 {count} 集\",\n \"totalWords\": \"总计 {count} 字\",\n \"autoSaved\": \"✓ 已自动保存\",\n \"reanalyze\": \"重新分析\",\n \"confirm\": \"确认完成\",\n \"saving\": \"保存中...\",\n \"episodeList\": \"剧集列表\",\n \"addEpisode\": \"添加剧集\",\n \"averageWords\": \"平均每集\",\n \"episodeContent\": \"剧集内容\",\n \"episodePlaceholder\": \"输入剧集标题...\",\n \"summaryPlaceholder\": \"输入剧情简介...\",\n \"newEpisode\": \"新剧集\",\n \"deleteEpisode\": \"删除剧集\",\n \"deleteConfirm\": {\n \"title\": \"确认删除\",\n \"message\": \"确定要删除「{title}」吗？\",\n \"cancel\": \"取消\",\n \"confirm\": \"确认删除\"\n },\n \"tip\": {\n \"title\": \"提示\",\n \"content\": \"你可以直接编辑标题、简介和内容。点击【确认完成】后，剧集将正式导入到项目中\"\n }\n },\n \"collapsePreview\": \"收起预览\",\n \"expandMore\": \"展开更多\",\n \"deleteFile\": \"删除文件\",\n \"fileTooLarge\": \"文件大小不能超过 10MB\",\n \"docNotSupported\": \"暂不支持 .doc 格式，请将文件另存为 .docx 或 .txt 格式后重试\",\n \"fileEmpty\": \"文件内容为空\",\n \"fileReadError\": \"文件读取失败，请确保文件格式正确\",\n \"uploadFirst\": \"请先上传文件或粘贴文本\",\n \"analyzeFailed\": \"分析失败\",\n \"saveFailed\": \"保存失败\",\n \"cancelConfirm\": \"确定要取消吗？已分析的剧集将被清空。\",\n \"deleteEpisode\": \"删除剧集\",\n \"upload\": {\n \"title\": \"上传原始素材\",\n \"subtitle\": \"AI 引擎已准备就绪，一键自动分集与格式化\",\n \"maxWords\": \"（最大支持 3 万字）\",\n \"textInput\": \"输入文本内容\",\n \"documentUpload\": \"上传完整文档\",\n \"placeholder\": \"在此处粘贴你的小说章节或剧本内容...\",\n \"filePlaceholder\": \"已上传文件模式\",\n \"clickUpload\": \"点击上传文档\",\n \"clearTextFirst\": \"请先清空左侧文本\",\n \"supportedFormats\": \"支持 Word, TXT 格式\",\n \"preview\": \"预览\",\n \"expandPreview\": \"展开更多\",\n \"collapsePreview\": \"收起预览\",\n \"deleteFile\": \"删除文件\",\n \"startAnalysis\": \"开始分析\",\n \"back\": \"返回\",\n \"words\": \"字\"\n },\n \"errors\": {\n \"fileTooLarge\": \"文件过大，请上传小于 10MB 的文件\",\n \"docNotSupported\": \"不支持 .doc 格式，请使用 Word 转换为 .docx\",\n \"fileEmpty\": \"文件内容为空\",\n \"fileReadError\": \"文件读取失败，请重试\",\n \"uploadFirst\": \"请先上传或粘贴内容\",\n \"analyzeFailed\": \"分析失败\",\n \"saveFailed\": \"保存失败\",\n \"analysisModelNotConfigured\": \"请先在设置页面配置分析模型\"\n },\n \"common\": {\n \"edit\": \"编辑\",\n \"delete\": \"删除\",\n \"save\": \"保存\",\n \"cancel\": \"取消\"\n },\n \"markerDetected\": {\n \"title\": \"检测到分集标记\",\n \"description\": \"检测到 {count} 个「{type}」格式的分集标记\",\n \"preview\": \"预览分集结果\",\n \"useMarker\": \"使用标识符分集\",\n \"useMarkerDesc\": \"快速、免费\",\n \"useAI\": \"使用 AI 智能分集\",\n \"useAIDesc\": \"智能分析、消耗积分\",\n \"cancel\": \"取消\",\n \"totalCount\": \"共 {count} 集\",\n \"markerTypes\": {\n \"episode\": \"第X集\",\n \"chapter\": \"第X章\",\n \"act\": \"第X幕\",\n \"scene\": \"X-Y【场景】\",\n \"numbered\": \"数字编号\",\n \"numberedEscaped\": \"数字编号(转义)\",\n \"numberedDirect\": \"数字+中文\",\n \"episodeEn\": \"Episode X\",\n \"chapterEn\": \"Chapter X\",\n \"boldNumber\": \"**数字**\",\n \"pureNumber\": \"纯数字\"\n }\n },\n \"globalAnalysis\": {\n \"title\": \"全局资产分析\",\n \"description\": \"一键提取全书的角色和场景，确保角色形象在所有剧集中保持一致\",\n \"startButton\": \"立即分析\",\n \"analyzing\": \"分析中...\",\n \"success\": \"全局分析完成：新增 {characters} 个角色，{locations} 个场景\",\n \"failed\": \"全局分析失败\",\n \"confirmAndAnalyze\": \"确认并开启全局分析\"\n }\n}"
},
{
"path": "messages/zh/stages.json",
"content": "{\n \"config\": \"1. 配置\",\n \"assets\": \"2. 资产分析\",\n \"storyboard\": \"3. 分镜编辑\",\n \"videos\": \"4. 视频生成\",\n \"voice\": \"5. 配音生成\"\n}"
},
{
"path": "messages/zh/storyboard.json",
"content": "{\n \"phases\": {\n \"planning\": \"规划分镜\",\n \"cinematography\": \"设计摄影\",\n \"acting\": \"设计演技\",\n \"detail\": \"补充细节\"\n },\n \"prompts\": {\n \"imagePrompt\": \"图片提示词\",\n \"aiInstruction\": \"AI修改指令\",\n \"supportReference\": \"(支持@引用资产库)\",\n \"instructionPlaceholder\": \"例如：把场景改为@医院_白天，人物改为@主角A\",\n \"selectAsset\": \"选择资产\",\n \"character\": \"人物\",\n \"location\": \"场景\",\n \"referencedAssets\": \"已引用资产：\",\n \"removeAsset\": \"移除此资产\",\n \"aiModify\": \"AI修改并生成图片\",\n \"aiModifying\": \"AI修改中...\",\n \"aiModifyTip\": \"点击后将自动保存提示词并生成新图片\",\n \"save\": \"保存\",\n \"currentPrompt\": \"当前提示词\",\n \"enterInstruction\": \"请输入修改指令\",\n \"modifyFailed\": \"操作失败: {error}\",\n \"updateFailed\": \"更新失败: {error}\",\n \"enterContinuation\": \"请输入续写内容\",\n \"appendTitle\": \"续写内容\",\n \"appendDescription\": \"输入新的SRT内容，系统会自动切分并生成新的镜头，追加到当前列表末尾\",\n \"appendSubmit\": \"续写并生成镜头\",\n \"appendSuccess\": \"续写成功！新镜头已追加到列表末尾\",\n \"appendFailed\": \"续写失败: {error}\",\n \"customStyle\": \"自定义风格\"\n },\n \"group\": {\n \"generating\": \"生成中...\",\n \"hasSynced\": \"✓ 已生成\",\n \"failed\": \"失败\",\n \"retry\": \"重试\",\n \"regenerate\": \"重新生成所有\",\n \"generateAll\": \"批量生成全部\",\n \"expand\": \"展开\",\n \"collapse\": \"收起\",\n \"addPanel\": \"添加镜头\",\n \"regenerating\": \"重新生成中...\",\n \"aiAnalyzing\": \"AI 分析中...\",\n \"regenerateText\": \"重新生成文字\",\n \"generateMissingImages\": \"生成该片段所有未有图片的镜头\",\n \"segment\": \"片段\",\n \"addAtStart\": \"在开头添加新分镜组\",\n \"insertHere\": \"在此插入新分镜组\"\n },\n \"header\": {\n \"title\": \"分镜编辑\",\n \"panels\": \"个镜头\",\n \"submit\": \"提交生成\",\n \"submitting\": \"提交中...\",\n \"storyboardPanel\": \"分镜面板\",\n \"segments\": \"个片段\",\n \"segmentsCount\": \"共 {count} 个片段，\",\n \"panelsCount\": \"{count} 个镜头\",\n \"generatingStatus\": \"({count} 个生成中)\",\n \"generateAllPanels\": \"生成所有镜头\",\n \"generatePendingPanels\": \"生成{count}个未有图片的镜头\",\n \"downloadAll\": \"下载全部\",\n \"downloading\": \"打包中...\",\n \"noImages\": \"没有可下载的图片\",\n \"downloadAllImages\": \"下载所有图片\",\n \"generateVideo\": \"生成视频 →\",\n \"back\": \"← 返回\",\n \"concurrencyLimit\": \"并发上限 {count}\"\n },\n \"panel\": {\n \"shotType\": \"景别：\",\n \"duration\": \"秒\",\n \"location\": \"场景：\",\n \"characters\": \"角色：\",\n \"description\": \"描述：\",\n \"text\": \"对应文本：\",\n \"regenerate\": \"重新生成\",\n \"delete\": \"删除\",\n \"insertBefore\": \"在此前插入\",\n \"insertAfter\": \"在此后插入\",\n \"moveUp\": \"上移\",\n \"moveDown\": \"下移\",\n \"plot\": \"剧情：\",\n \"summary\": \"总结：\",\n \"pov\": \"视角：\",\n \"focus\": \"焦点：\",\n \"mode\": \"模式：\",\n \"shot\": \"镜头\",\n \"segment\": \"片段\",\n \"stylePrompt\": \"画风/提示词\",\n \"shotMode\": \"景别/模式\",\n \"regenerateImage\": \"重新生成图片\",\n \"generateImage\": \"生成图片\",\n \"cardView\": \"卡片视图\",\n \"tableView\": \"表格视图\",\n \"shotTypeLabel\": \"镜头类型\",\n \"cameraMove\": \"镜头运动\",\n \"sourceText\": \"对应原文\",\n \"sceneDescription\": \"画面描述\",\n \"videoPrompt\": \"视频提示词\",\n \"videoPromptHint\": \"建议描述主体动作、环境、镜头语言\",\n \"locationLabel\": \"场景\",\n \"editLocation\": \"编辑场景\",\n \"characterLabel\": \"角色\",\n \"characterLabelWithCount\": \"角色 ({count})\",\n \"editCharacter\": \"编辑角色\",\n \"select\": \"+ 选择\",\n \"add\": \"+ 添加\",\n \"noLocation\": \"未选择场景\",\n \"locationNotEdited\": \"暂未编辑场景\",\n \"noCharacters\": \"未选择角色\",\n \"charactersNotEdited\": \"暂未编辑角色\",\n \"shotTypePlaceholder\": \"俯拍中景...\",\n \"cameraMovePlaceholder\": \"缓推、固定...\",\n \"videoPromptPlaceholder\": \"用于视频生成的提示词...\",\n \"sceneDescriptionPlaceholder\": \"描述画面主体、构图、光线、情绪\",\n \"selectCharacter\": \"选择角色\",\n \"selectLocation\": \"选择场景\",\n \"noCharacterAssets\": \"暂无角色资产\",\n \"noLocationAssets\": \"暂无场景资产\",\n \"selected\": \"已选择\",\n \"defaultAppearance\": \"初始形象\",\n \"newPanelDescription\": \"新镜头描述\",\n \"noShotType\": \"未设置镜头\"\n },\n \"image\": {\n \"generating\": \"生成中...\",\n \"regenerate\": \"重新生成\",\n \"edit\": \"编辑\",\n \"editImage\": \"修图\",\n \"candidate\": \"候选图\",\n \"selectCandidate\": \"选择候选\",\n \"variants\": \"变体\",\n \"generateVariants\": \"生成变体\",\n \"forceRegenerate\": \"强制重新生成\",\n \"failed\": \"生成失败\",\n \"clickToPreview\": \"点击放大预览\",\n \"enlargePreview\": \"放大预览\",\n \"candidateCount\": \"候选 {count}\",\n \"candidateGenerating\": \"{count} 张生成中\",\n \"selectingCandidate\": \"选择候选图中...\",\n \"confirmCandidate\": \"确认选择\",\n \"cancelSelection\": \"取消选择\",\n \"noValidCandidates\": \"暂无有效候选图\",\n \"selectCount\": \"选择生成数量\",\n \"generateMultiple\": \"生成多张候选\",\n \"generateCount\": \"生成 {count} 张\",\n \"generateCountSuffix\": \"张\",\n \"undoShort\": \"返回\"\n },\n \"candidate\": {\n \"title\": \"选择候选图\",\n \"select\": \"选择\",\n \"cancel\": \"取消\",\n \"noImages\": \"暂无候选图\",\n \"original\": \"原图\"\n },\n \"variant\": {\n \"title\": \"图片变体\",\n \"generate\": \"生成变体\",\n \"select\": \"使用此图\",\n \"close\": \"关闭\",\n \"shotTitle\": \"镜头变体 - 基于 #{number}\",\n \"originalDescription\": \"原镜头描述\",\n \"noDescription\": \"无描述\",\n \"noImage\": \"无图片\",\n \"shotNum\": \"镜头 {number}\",\n \"aiRecommend\": \"AI 推荐变体\",\n \"reanalyze\": \"重新分析\",\n \"shotType\": \"景别:\",\n \"cameraMove\": \"运镜:\",\n \"generating\": \"生成中\",\n \"clickToAnalyze\": \"点击重新分析获取 AI 推荐\",\n \"customInstruction\": \"或自定义指令\",\n \"customPlaceholder\": \"输入你想要的镜头效果，如：改为反打视角，聚焦另一个角色的表情...\",\n \"includeCharacter\": \"引用角色形象\",\n \"includeLocation\": \"引用场景图\",\n \"customVariant\": \"自定义变体\",\n \"defaultShotType\": \"中景\",\n \"defaultCameraMove\": \"固定\",\n \"useCustomGenerate\": \"使用自定义生成\",\n \"analyzeFailed\": \"分析失败\",\n \"creativeScore\": \"创意 {score}/5\"\n },\n \"insert\": {\n \"title\": \"插入新镜头\",\n \"position\": \"插入位置\",\n \"before\": \"在第 {number} 镜头前\",\n \"after\": \"在第 {number} 镜头后\",\n \"content\": \"镜头内容\",\n \"shotType\": \"景别\",\n \"location\": \"场景\",\n \"characters\": \"角色\",\n \"description\": \"描述\",\n \"text\": \"对应文本\",\n \"placeholder\": {\n \"shotType\": \"选择景别...\",\n \"location\": \"输入场景...\",\n \"characters\": \"输入角色，用逗号分隔\",\n \"description\": \"描述画面内容...\",\n \"text\": \"对应的剧本文本...\"\n },\n \"insert\": \"插入\",\n \"cancel\": \"取消\"\n },\n \"common\": {\n \"actions\": \"操作\",\n \"add\": \"添加\",\n \"cancel\": \"取消\",\n \"confirm\": \"确认\",\n \"copy\": \"复制\",\n \"delete\": \"删除\",\n \"download\": \"下载\",\n \"edit\": \"编辑\",\n \"generate\": \"生成\",\n \"loading\": \"加载中...\",\n \"none\": \"无\",\n \"unknownError\": \"未知错误\",\n \"preview\": \"预览\",\n \"refresh\": \"刷新\",\n \"regenerate\": \"重新生成\",\n \"deleting\": \"删除中\",\n \"editing\": \"编辑中\",\n \"saving\": \"保存中...\",\n \"saveFailed\": \"保存失败，修改尚未同步\",\n \"retrySave\": \"重试保存\",\n \"save\": \"保存\",\n \"status\": \"状态\",\n \"submitFailed\": \"提交失败\",\n \"upload\": \"上传\"\n },\n \"confirm\": {\n \"deletePanel\": \"确定要删除这个镜头吗？删除后无法恢复。\",\n \"deleteGroup\": \"确定要删除这整组分镜吗？\\n\\n这将删除该片段下的所有 {count} 个镜头，此操作不可撤销！\"\n },\n \"messages\": {\n \"episodeNotFound\": \"没有找到剧集信息\",\n \"downloadFailed\": \"下载失败: {error}\",\n \"panelNotFound\": \"未找到镜头信息\",\n \"modifyFailed\": \"修改失败: {error}\",\n \"selectCandidateFailed\": \"选择失败: {error}\",\n \"insertPanelFailed\": \"插入分镜失败: {error}\",\n \"addPanelFailed\": \"添加分镜失败: {error}\",\n \"deletePanelFailed\": \"删除失败: {error}\",\n \"deleteGroupFailed\": \"删除分镜组失败: {error}\",\n \"regenerateGroupFailed\": \"重新生成分镜失败: {error}\",\n \"addGroupFailed\": \"添加分镜组失败: {error}\",\n \"moveGroupFailed\": \"移动分镜组失败: {error}\",\n \"batchGenerateCompleted\": \"批量生成完成:\\n成功: {succeeded}\\n失败: {failed}\\n\\n部分错误: {errors}\",\n \"batchGenerateFailed\": \"批量生成失败: {error}\"\n },\n \"canvas\": {\n \"emptyTitle\": \"暂无分镜数据\",\n \"emptyDescription\": \"请先生成Clips和文字分镜，或点击上方按钮添加分镜组\"\n },\n \"imageEdit\": {\n \"title\": \"编辑分镜\",\n \"subtitle\": \"输入修改指令，可选择上传参考图片和资产\",\n \"promptPlaceholder\": \"描述你想要修改的内容，例如：改变背景颜色、调整人物表情...\",\n \"referenceImagesLabel\": \"参考图片\",\n \"referenceImagesHint\": \"(可选，支持粘贴)\",\n \"start\": \"开始编辑\",\n \"selectAsset\": \"选择资产\",\n \"selectedAssetsLabel\": \"参考资产\",\n \"selectedAssetsCount\": \"{count}个\",\n \"addAsset\": \"添加资产\",\n \"noAssets\": \"暂无资产，点击“添加资产”选择\"\n },\n \"screenplay\": {\n \"tabs\": {\n \"formatted\": \"剧本格式\",\n \"original\": \"原文\"\n },\n \"scene\": \"场景 {number}\",\n \"characters\": \"出场角色\",\n \"voiceover\": \"旁白\",\n \"parseFailedTitle\": \"剧本格式解析失败\",\n \"parseFailedDescription\": \"请查看原文内容\"\n },\n \"assets\": {\n \"character\": {\n \"confirming\": \"确认中...\",\n \"editing\": \"编辑中...\"\n },\n \"image\": {\n \"undo\": \"撤销到上一版本\"\n },\n \"location\": {\n \"generateImage\": \"生成图片\"\n },\n \"stage\": {\n \"analyzing\": \"分析中...\"\n }\n },\n \"video\": {\n \"toolbar\": {\n \"showPending\": \"待生成\"\n },\n \"panelCard\": {\n \"forceRegenerate\": \"强制重新生成（卡住时使用）\"\n }\n },\n \"smartImport\": {\n \"errors\": {\n \"analyzeFailed\": \"分析失败\"\n },\n \"preview\": {\n \"reanalyze\": \"重新分析\"\n },\n \"smartImport\": {\n \"recommended\": \"推荐\"\n }\n },\n \"aiData\": {\n \"title\": \"AI数据编辑器\",\n \"subtitle\": \"Panel {number} - 发送给图片生成AI的完整数据\",\n \"basicData\": \"分镜基础数据\",\n \"shotType\": \"镜头类型\",\n \"cameraMove\": \"镜头运动\",\n \"shotTypePlaceholder\": \"仰拍、全景、平视、中景...\",\n \"cameraMovePlaceholder\": \"缓推、固定、跟随...\",\n \"scene\": \"场景（只读）\",\n \"notSelected\": \"未选择\",\n \"summary\": \"场景总结\",\n \"characters\": \"角色（只读）\",\n \"plot\": \"剧情\",\n \"summarize\": \"总结\",\n \"visualDescription\": \"视觉描述\",\n \"videoPrompt\": \"视频提示词\",\n \"negativePrompt\": \"负面提示词\",\n \"save\": \"保存\",\n \"cancel\": \"取消\",\n \"lightingDirection\": \"光照方向\",\n \"lightingQuality\": \"光照质感\",\n \"depthOfField\": \"景深\",\n \"colorTone\": \"色调\",\n \"characterPosition\": \"角色位置规则\",\n \"position\": \"位置\",\n \"posture\": \"姿势\",\n \"facing\": \"朝向\",\n \"photographyRules\": \"摄影规则 (photography_rules)\",\n \"viewData\": \"查看数据\",\n \"jsonPreview\": \"JSON 预览\",\n \"actingNotes\": \"演技指导 (acting_notes)\",\n \"actingTitle\": \"演技指导\",\n \"actingDescription\": \"表演指令\",\n \"noActingData\": \"无演技数据\"\n },\n \"insertModal\": {\n \"insertBetween\": \"在 #{before} 和 #{after} 之间插入\",\n \"panel\": \"镜头\",\n \"noImage\": \"无图片\",\n \"insertAtEnd\": \"末尾\",\n \"aiAnalyze\": \"AI 自动分析\",\n \"analyzing\": \"AI 分析中...\",\n \"insert\": \"插入\",\n \"inserting\": \"插入中...\",\n \"placeholder\": \"可选：输入补充说明，如添加一个反应镜头...\"\n },\n \"panelActions\": {\n \"insertPanel\": \"插入分镜\",\n \"panelVariant\": \"镜头变体\",\n \"insertHere\": \"在此处插入分镜\",\n \"generateVariant\": \"基于此镜头生成变体\",\n \"needImage\": \"需要先生成图片\",\n \"deleteShot\": \"删除镜头\",\n \"pasteSrtPlaceholder\": \"粘贴新的SRT内容...\"\n },\n \"firstLastFrame\": {\n \"placeholder\": \"输入首尾帧视频提示词...\",\n \"modelTitle\": \"首尾帧模型\"\n }\n}\n"
},
{
"path": "messages/zh/video.json",
"content": "{\n \"panelCard\": {\n \"play\": \"播放\",\n \"pause\": \"暂停\",\n \"retry\": \"重试\",\n \"regenerate\": \"重新生成\",\n \"download\": \"下载\",\n \"edit\": \"编辑\",\n \"save\": \"保存\",\n \"cancel\": \"取消\",\n \"generating\": \"生成中...\",\n \"failed\": \"失败\",\n \"lipSync\": \"口型同步\",\n \"lipSyncVideo\": \"口型同步视频\",\n \"lipSyncLabel\": \"口型同步\",\n \"lipSyncTitle\": \"口型同步\",\n \"original\": \"原始\",\n \"synced\": \"同步\",\n \"videoFixed\": \"✓ 视频\",\n \"imagePreview\": \"图片预览\",\n \"playVoice\": \"播放配音\",\n \"stopVoice\": \"停止\",\n \"noVoice\": \"暂无配音\",\n \"forceRegenerate\": \"强制重新生成（卡住时使用）\",\n \"regenerateVideo\": \"重新生成视频\",\n \"lipSyncStatus\": \"口型同步中...\",\n \"lipSyncInProgress\": \"正在进行口型同步...\",\n \"lipSyncMayTakeMinutes\": \"这可能需要几分钟时间\",\n \"audioEnabled\": \"音频已开启\",\n \"audioDisabled\": \"音频已关闭\",\n \"isSynced\": \"(已同步)\",\n \"needVideo\": \"(请先生成视频)\",\n \"needAudio\": \"(请先生成音频)\",\n \"generateAudio\": \"生成音频\",\n \"regenerateLipSync\": \"重新生成口型同步\",\n \"editPrompt\": \"编辑提示词\",\n \"clickToEditPrompt\": \"点击编辑提示词...\",\n \"shot\": \"镜头 {number}\",\n \"unknownShotType\": \"未知景别\",\n \"correspondingText\": \"对应原文\",\n \"generateVideo\": \"生成视频\",\n \"selectModel\": \"选择视频模型\",\n \"selectVoice\": \"选择要使用的配音：\",\n \"willAutoPad\": \"(将自动填充)\",\n \"autoPadding\": \"填充\",\n \"redo\": \"重新\",\n \"generatingAudio\": \"生成中\",\n \"error\": {\n \"audioFailed\": \"生成音频失败\"\n },\n \"batchMode\": \"批量模式\",\n \"batchModeDesc\": \"离线推理，价格便宜50%，24小时内完成\",\n \"batchModeEnabled\": \"已开启批量模式\",\n \"batchModeDisabled\": \"批量模式已关闭\"\n },\n \"promptModal\": {\n \"title\": \"编辑镜头 #{number} 视频提示词\",\n \"shotType\": \"景别：\",\n \"duration\": \"秒\",\n \"location\": \"场景：\",\n \"locationUnknown\": \"未知\",\n \"characters\": \"角色：\",\n \"charactersNone\": \"无\",\n \"description\": \"描述：\",\n \"text\": \"对应文本：\",\n \"promptLabel\": \"视频提示词\",\n \"placeholder\": \"输入视频提示词...\",\n \"tip\": \"提示：视频模型不识别角色名字，请用外貌特征描述，如\\\"黑发蓝眸的年轻男子\\\"而非\\\"Victor\\\"\",\n \"save\": \"保存\",\n \"cancel\": \"取消\"\n },\n \"toolbar\": {\n \"title\": \"成片生成\",\n \"filter\": \"筛选\",\n \"viewAll\": \"查看全部\",\n \"showGenerated\": \"已生成\",\n \"showPending\": \"待生成\",\n \"showFailed\": \"失败\",\n \"totalShots\": \"共 {count} 个镜头\",\n \"generatingShots\": \"{count} 个生成中\",\n \"completedShots\": \"{count} 个已生成\",\n \"failedShots\": \"{count} 个失败\",\n \"generateAll\": \"生成所有视频\",\n \"batchConfigTitle\": \"批量生成参数\",\n \"batchConfigDesc\": \"先选择模型与参数，再一键生成所有视频\",\n \"confirmGenerateAll\": \"确认并生成全部\",\n \"confirming\": \"提交中...\",\n \"noVideos\": \"没有可下载的视频\",\n \"downloadCount\": \"下载 {count} 个视频\",\n \"packing\": \"打包中...\",\n \"downloadAll\": \"下载全部\",\n \"enterEditor\": \"进入视频剪辑器\",\n \"enterEdit\": \"进入剪辑\",\n \"back\": \"返回\"\n },\n \"stage\": {\n \"title\": \"视频生成\",\n \"generateAll\": \"批量生成全部\",\n \"regenerateFailed\": \"重试失败项\",\n \"downloadAll\": \"下载全部视频\",\n \"enterEditor\": \"进入剪辑器\",\n \"lipSyncStatus\": \"口型同步中...\",\n \"hasSynced\": \"✓ 已生成\",\n \"generating\": \"生成中...\",\n \"downloading\": \"下载中...\",\n \"downloadProgress\": \"正在准备视频文件... {current}/{total}\",\n \"noVideos\": \"暂无已生成的视频\",\n \"scrollTo\": \"跳转到镜头\",\n \"error\": {\n \"saveFailed\": \"保存视频提示词失败\",\n \"lipSyncFailed\": \"口型同步失败\",\n \"fetchVideosFailed\": \"获取视频列表失败\"\n },\n \"downloadFailed\": \"下载失败\",\n \"unknownError\": \"未知错误\"\n },\n \"firstLastFrame\": {\n \"title\": \"首尾帧设置\",\n \"firstFrame\": \"首帧\",\n \"lastFrame\": \"尾帧\",\n \"range\": \"镜头 {from} → 镜头 {to}\",\n \"link\": \"链接\",\n \"unlink\": \"解除链接\",\n \"unlinkAction\": \"取消链接\",\n \"asLastFrameFor\": \"作为镜头 {number} 的尾帧\",\n \"asFirstFrameFor\": \"作为镜头 {number} 的首帧\",\n \"customPrompt\": \"自定义提示词\",\n \"promptPlaceholder\": \"输入首尾帧视频提示词...\",\n \"useDefault\": \"使用默认\",\n \"generate\": \"生成首尾帧视频\",\n \"generated\": \"首尾帧视频已生成\",\n \"model\": \"模型\",\n \"withAudio\": \"包含音频\",\n \"audioOn\": \"开\",\n \"audioOff\": \"关\",\n \"linkToNext\": \"链接到下一镜头（首尾帧）\",\n \"asLastFrame\": \"作为镜头 {number} 的尾帧\",\n \"thenTransitionTo\": \"然后镜头转换到\"\n },\n \"editor\": {\n \"alert\": {\n \"saveSuccess\": \"保存成功\",\n \"saveFailed\": \"保存失败\",\n \"exportStarted\": \"导出任务已开始，请稍候...\",\n \"exportFailed\": \"导出失败\"\n },\n \"toolbar\": {\n \"back\": \"← 返回\",\n \"saveDirty\": \"保存 *\",\n \"saved\": \"已保存\",\n \"export\": \"导出视频\"\n },\n \"left\": {\n \"title\": \"素材库\",\n \"description\": \"从视频阶段导入的片段将显示在这里\"\n },\n \"right\": {\n \"title\": \"属性\",\n \"clipLabel\": \"片段:\",\n \"clipFallback\": \"片段 {index}\",\n \"durationLabel\": \"时长:\",\n \"transitionLabel\": \"转场到下一片段\",\n \"deleteConfirm\": \"确定删除此片段？\",\n \"deleteClip\": \"删除片段\",\n \"selectClipHint\": \"选择一个片段查看属性\"\n },\n \"preview\": {\n \"emptyStartEditing\": \"添加素材开始编辑\"\n },\n \"timeline\": {\n \"zoomLabel\": \"缩放:\",\n \"videoTrack\": \"视频\",\n \"emptyHint\": \"从素材库拖拽视频片段到这里\",\n \"audioTrack\": \"配音\",\n \"audioBadge\": \"配\"\n },\n \"transition\": {\n \"title\": \"转场效果\",\n \"duration\": \"持续时间\",\n \"options\": {\n \"none\": \"无\",\n \"dissolve\": \"溶解\",\n \"fade\": \"淡入淡出\",\n \"slide\": \"滑动\"\n }\n }\n },\n \"errors\": {\n \"unknownError\": \"未知错误\"\n },\n \"capability\": {\n \"generationMode\": \"生成模式\",\n \"generateAudio\": \"生成音频\",\n \"duration\": \"时长\",\n \"fps\": \"帧率\",\n \"resolution\": \"分辨率\",\n \"aspectRatio\": \"画幅比例\",\n \"reasoningEffort\": \"推理强度\",\n \"voice\": \"音色\",\n \"rate\": \"语速\",\n \"mode\": \"模式\"\n },\n \"unit\": {\n \"second\": \"秒\",\n \"frame\": \"帧\"\n },\n \"common\": {\n \"generate\": \"生成\"\n }\n}"
},
{
"path": "messages/zh/voice.json",
"content": "{\n \"title\": \"台词配音\",\n \"linesCount\": \"共 {count} 条台词，\",\n \"audioGeneratedCount\": \"{count} 条已生成音频\",\n \"emotionPrompt\": \"情绪提示词\",\n \"emotionPromptTip\": \"(不填则使用台词自参考)\",\n \"emotionPlaceholder\": \"如：laugh，仅支持英文...\",\n \"emotionStrength\": \"情绪强度\",\n \"flat\": \"平淡\",\n \"intense\": \"强烈\",\n \"generating\": \"生成中...\",\n \"generateVoice\": \"生成语音\",\n \"toolbar\": {\n \"back\": \"← 返回\",\n \"analyzeLines\": \"分析台词\",\n \"addLine\": \"＋ 添加语音\",\n \"generateAll\": \"一键生成所有配音\",\n \"downloadAll\": \"下载配音\",\n \"generatingCount\": \"生成中 ({count})\",\n \"packing\": \"打包中...\",\n \"stats\": \"共 {total} 条台词 | 已设置音色 {withVoice} 条 | 已生成配音 {withAudio} 条\",\n \"noDownload\": \"没有可下载的配音\",\n \"downloadCount\": \"下载 {count} 条配音\",\n \"uploadReferenceHint\": \"请先在资产库为所有角色上传参考音频\"\n },\n \"speakerVoice\": {\n \"title\": \"发言人音色状态\",\n \"hint\": \"请在资产库为角色上传参考音频\",\n \"linesCount\": \"{count} 条台词\",\n \"noVoice\": \"无参考音色\",\n \"configured\": \"✓ 已设置\",\n \"playVoice\": \"播放当前音色\",\n \"aiDesign\": \"AI设计声音\",\n \"aiDesignVoice\": \"AI 设计音色\",\n \"redesign\": \"使用 AI 重新设计声音\",\n \"uploadAudio\": \"上传音频\",\n \"uploading\": \"上传中\",\n \"upload\": \"上传\",\n \"microsoftVoice\": \"微软语音\",\n \"microsoft\": \"微软\",\n \"maleVoices\": \"男声\",\n \"femaleVoices\": \"女声\",\n \"openAssetLibrary\": \"资产库\",\n \"configuredStatus\": \"已设置音色\",\n \"pendingStatus\": \"待设置音色\",\n \"voiceSettings\": \"音色设置\",\n \"inlineLabel\": \"内联\"\n },\n \"inlineBinding\": {\n \"title\": \"为「{speaker}」设置音色\",\n \"description\": \"该发言人不在资产库中，请选择一种方式为其设置参考音色\",\n \"selectFromLibrary\": \"从音色库选择\",\n \"selectFromLibraryDesc\": \"选择已有的全局音色\",\n \"uploadAudio\": \"上传参考音频\",\n \"uploadAudioDesc\": \"上传 MP3、WAV 等音频文件作为参考音色\",\n \"uploadQwenHint\": \"上传的音色后续只可使用 IndexTTS 合成，不可用于 QwenTTS。QwenTTS 必须使用 AI 设计音色。\",\n \"aiDesign\": \"AI 设计音色\",\n \"aiDesignDesc\": \"使用 AI 生成专属参考音色\"\n },\n \"embedded\": {\n \"linesStats\": \"{total} 条台词 · {audio} 已生成\",\n \"reanalyze\": \"重新分析\",\n \"analyzeLines\": \"分析台词\",\n \"reanalyzeHint\": \"重新分析台词并更新镜头匹配\",\n \"analyzeHint\": \"从原文中提取台词\",\n \"downloadVoice\": \"下载配音\",\n \"generateAllVoice\": \"生成全部配音\",\n \"pendingCount\": \"({count} 条待生成)\",\n \"generatingProgress\": \"生成中 ({current}/{total})\",\n \"generatingHint\": \"正在生成中...\",\n \"noVoiceHint\": \"请先在上方为所有角色设置音色\",\n \"noLinesHint\": \"没有台词可生成\",\n \"allDoneHint\": \"所有台词已生成完成\",\n \"generateHint\": \"点击生成 {count} 条待生成的配音\",\n \"addLine\": \"＋ 添加语音\",\n \"speakerVoiceStatus\": \"角色音色状态\",\n \"speakersCount\": \"{count} 个\",\n \"listen\": \"试听\",\n \"listenVoice\": \"试听音色\",\n \"reset\": \"重设\",\n \"resetDesign\": \"重新设计\",\n \"aiDesign\": \"AI设计\",\n \"assetLibrary\": \"资产库\"\n },\n \"lineCard\": {\n \"generatingVoice\": \"生成中\",\n \"speaker\": \"发言人\",\n \"speakerPlaceholder\": \"发言人名称\",\n \"content\": \"台词内容\",\n \"contentPlaceholder\": \"台词内容\",\n \"emotionConfigured\": \"情绪已设置\",\n \"emotionSettings\": \"情绪设置\",\n \"voiceConfigured\": \"✓ 已设置\",\n \"needVoice\": \"请在上方设置音色\",\n \"locatePanel\": \"定位到绑定镜头\",\n \"locateVideo\": \"定位视频\",\n \"play\": \"播放\",\n \"pause\": \"暂停\",\n \"locatePanelCta\": \"定位到镜头 {index}\",\n \"editLine\": \"编辑台词\",\n \"deleteLine\": \"删除台词\",\n \"deleteAudio\": \"删除配音\"\n },\n \"lineEditor\": {\n \"addTitle\": \"添加语音\",\n \"editTitle\": \"编辑语音\",\n \"contentLabel\": \"台词内容\",\n \"contentPlaceholder\": \"请输入台词内容\",\n \"speakerLabel\": \"发言人\",\n \"speakerPlaceholder\": \"请输入发言人名称\",\n \"selectSpeaker\": \"请选择发言人\",\n \"noSpeakerOptions\": \"当前项目暂无可选发言人，请先分析台词生成发言人\",\n \"bindPanelLabel\": \"绑定镜头\",\n \"unboundPanel\": \"未绑定镜头\",\n \"panelLabel\": \"镜头 {index}\",\n \"saveAdd\": \"添加语音\",\n \"saveEdit\": \"保存修改\"\n },\n \"empty\": {\n \"title\": \"暂无台词数据\",\n \"description\": \"从剧本中提取台词和发言人\",\n \"analyzeButton\": \"分析台词\",\n \"hint\": \"请先在资产库为角色上传参考音频\"\n },\n \"confirm\": {\n \"deleteLine\": \"确定要删除这条台词吗？\\n\\n\\\"{content}\\\"\\n\\n此操作不可撤销。\",\n \"deleteAudio\": \"确定要删除这条台词的配音吗？\\n\\n\\\"{content}\\\"\\n\\n此操作不可撤销。\"\n },\n \"errors\": {\n \"saveFailed\": \"保存失败\",\n \"analyzeFailed\": \"分析台词失败\",\n \"generateFailed\": \"生成配音失败\",\n \"batchFailed\": \"批量生成失败\",\n \"downloadFailed\": \"下载失败\",\n \"deleteFailed\": \"删除失败\",\n \"addFailed\": \"添加语音失败\",\n \"invalidLineInput\": \"台词内容和发言人不能为空\",\n \"bindFailed\": \"绑定镜头失败\",\n \"deleteAudioFailed\": \"删除配音失败\",\n \"uploadFailed\": \"上传音频失败\",\n \"voiceDesignFailed\": \"保存声音设计失败\",\n \"emotionSaveFailed\": \"保存情绪设置失败\",\n \"voiceGenerateFailed\": \"生成音频失败\"\n },\n \"alerts\": {\n \"insufficientBalance\": \"余额不足\",\n \"insufficientBalanceMsg\": \"账户余额不足，请充值后继续使用\",\n \"noLinesToGenerate\": \"没有需要生成的台词（请先为角色上传参考音频）\",\n \"generateComplete\": \"生成完成：{success}/{total} 条成功\",\n \"generateFailed\": \"{count} 条失败\",\n \"speakerVoiceSet\": \"已为 {speaker} 生成参考音频\",\n \"speakerVoiceUploaded\": \"已为 {speaker} 上传参考音频\",\n \"voiceDesignSet\": \"已为 {speaker} 设置 AI 设计的声音\"\n },\n \"common\": {\n \"loading\": \"加载中...\",\n \"save\": \"保存\",\n \"cancel\": \"取消\",\n \"cancelling\": \"取消中...\",\n \"upload\": \"上传\",\n \"download\": \"下载\",\n \"generate\": \"生成\",\n \"regenerate\": \"重新生成\"\n },\n \"assets\": {\n \"image\": {\n \"uploadFailed\": \"上传失败\"\n },\n \"stage\": {\n \"analyzing\": \"分析中...\"\n }\n },\n \"smartImport\": {\n \"errors\": {\n \"analyzeFailed\": \"分析失败\"\n }\n },\n \"video\": {\n \"panelCard\": {\n \"play\": \"播放\"\n }\n },\n \"tts\": {\n \"generatedAudio\": \"生成的音频\",\n \"browserNotSupport\": \"您的浏览器不支持音频播放\",\n \"audioDuration\": \"音频时长：\",\n \"subtitleCount\": \"字幕条数：\",\n \"noAudio\": \"暂无音频\",\n \"srtPreview\": \"SRT字幕预览\",\n \"noSubtitle\": \"暂无字幕\",\n \"stats\": \"生成统计\",\n \"minute\": \"分\",\n \"second\": \"秒\",\n \"items\": \"条\",\n \"completed\": \"✓ 已完成\",\n \"regenerating\": \"重新生成中...\",\n \"regenerateTTS\": \"重新生成TTS\",\n \"nextStep\": \"下一步: 分析资产\",\n \"readyTip\": \"点击进入资产分析阶段\",\n \"needGenerate\": \"请先生成TTS音频\"\n },\n \"voiceCreate\": {\n \"aiDesignMode\": \"AI 设计音色\",\n \"uploadMode\": \"上传音频\",\n \"dropOrClick\": \"拖放文件或点击选择\",\n \"supportedFormats\": \"支持 MP3、WAV、OGG、M4A、AAC 格式\",\n \"invalidFileType\": \"不支持的文件格式，请上传音频文件\",\n \"fileTooLarge\": \"文件过大，最大支持 50MB\",\n \"previewAudio\": \"试听音频\",\n \"uploading\": \"上传中...\",\n \"uploadFailed\": \"上传失败\",\n \"uploadSuccess\": \"上传成功\"\n },\n \"voiceDesign\": {\n \"presets\": {\n \"maleBroadcaster\": \"男播音\",\n \"gentleFemale\": \"温柔女\",\n \"matureMale\": \"成熟男\",\n \"livelyFemale\": \"活泼女\",\n \"intellectualFemale\": \"知性女\",\n \"narrator\": \"旁白\"\n },\n \"presetsPrompts\": {\n \"maleBroadcaster\": \"沉稳的中年男性播音员，音色低沉浑厚，语速平稳，吐字清晰\",\n \"gentleFemale\": \"温柔甜美的年轻女性，声音清脆悦耳，语调轻柔\",\n \"matureMale\": \"成熟稳重的男性，声音富有磁性和感染力\",\n \"livelyFemale\": \"活泼开朗的少女，声音甜美可爱，充满活力\",\n \"intellectualFemale\": \"知性优雅的女性，声音清晰悦耳，语调平和\",\n \"narrator\": \"富有感情的叙述者，声音温暖有故事感\"\n },\n \"defaultPreviewText\": \"你好，很高兴认识你。这是AI为你专属设计的声音，让我来为你展示它的特点。无论是温柔的对话，还是激动的讲述，我都能完美呈现。希望你喜欢这个声音，让我们一起创造精彩的内容吧。\",\n \"pleaseSelectStyle\": \"请输入或选择声音风格\",\n \"designVoiceFor\": \"为「{speaker}」设计AI声音\",\n \"hasExistingVoice\": \"已有声音\",\n \"selectStyle\": \"选择声音风格：\",\n \"orCustomDescription\": \"或自定义描述：\",\n \"describePlaceholder\": \"描述声音特征：年龄、性别、音色、语调...\",\n \"generateSchemesPrefix\": \"生成\",\n \"generateSchemesSuffix\": \"个声音方案\",\n \"schemeCountAriaLabel\": \"生成声音方案数量\",\n \"editPreviewText\": \"修改预览文本\",\n \"generating3Schemes\": \"正在生成 3 个声音方案...\",\n \"estimatedTime\": \"预计 15-30 秒\",\n \"selectScheme\": \"选择声音方案：\",\n \"schemeN\": \"方案 {n}\",\n \"regenerate\": \"重新生成\",\n \"confirmUse\": \"✓ 确认使用\",\n \"confirmReplace\": \"确认替换声音？\",\n \"replaceWarning\": \"的原有声音，不可撤回\",\n \"confirmReplaceBtn\": \"确认替换\",\n \"noVoiceGenerated\": \"未能生成任何声音\",\n \"generationError\": \"生成声音失败\",\n \"generateFailed\": \"生成第 {n} 个声音失败\",\n \"preview\": \"试听\",\n \"playing\": \"播放中\"\n }\n}\n"
},
{
"path": "messages/zh/workspace.json",
"content": "{\n \"title\": \"我的项目\",\n \"subtitle\": \"管理您的AI动漫制作项目\",\n \"newProject\": \"新建项目\",\n \"searchPlaceholder\": \"搜索项目名称或描述...\",\n \"searchButton\": \"搜索\",\n \"clearButton\": \"清除\",\n \"updatedAt\": \"更新于\",\n \"noProjects\": \"还没有项目\",\n \"noProjectsDesc\": \"创建您的第一个AI动漫制作项目\",\n \"noResults\": \"没有找到匹配的项目\",\n \"noResultsDesc\": \"尝试使用不同的搜索词\",\n \"createProject\": \"新建项目\",\n \"editProject\": \"编辑项目\",\n \"deleteProject\": \"删除项目\",\n \"deleteConfirm\": \"确定要删除项目\\\"{name}\\\"吗？此操作无法撤销。\",\n \"projectName\": \"项目名称\",\n \"projectNamePlaceholder\": \"输入项目名称\",\n \"projectDescription\": \"项目描述（可选）\",\n \"projectDescriptionPlaceholder\": \"输入项目描述\",\n \"creating\": \"创建中...\",\n \"saving\": \"保存中...\",\n \"createFailed\": \"创建项目失败\",\n \"analysisModelRequiredAfterCreate\": \"项目已创建。请先前往个人设置配置默认模型（至少设置分析模型），否则无法使用。\",\n \"updateFailed\": \"更新项目失败\",\n \"deleteFailed\": \"删除项目失败\",\n \"totalProjects\": \"共 {count} 个项目\",\n \"statsEpisodes\": \"章节数\",\n \"statsImages\": \"图片数\",\n \"statsVideos\": \"视频数\",\n \"noContent\": \"暂无内容\",\n \"modelNotConfigured\": {\n \"before\": \"检测到尚未配置模型，请先前往\",\n \"link\": \"设置中心\",\n \"after\": \"配置模型，或在创建项目后于项目配置中自定义。\"\n }\n}"
},
{
"path": "messages/zh/workspaceDetail.json",
"content": "{\n \"globalAssets\": \"全局资产\",\n \"createFailed\": \"创建失败\",\n \"deleteFailed\": \"删除失败\",\n \"renameFailed\": \"重命名失败\",\n \"refreshFailed\": \"刷新失败\",\n \"projectNotFound\": \"项目不存在\",\n \"backToWorkspace\": \"返回工作区\",\n \"episode\": \"剧集\",\n \"modelSetup\": {\n \"title\": \"开始创作前，请先配置默认模型\",\n \"description\": \"当前账号还没有默认分析模型，智能分集和后续 AI 流程无法使用。请先完成默认模型配置。\",\n \"configureNow\": \"立即配置分析模型\",\n \"goProfile\": \"前往个人设置\",\n \"modalTitle\": \"配置默认分析模型\",\n \"modalDescription\": \"请选择一个可用的文本模型作为默认分析模型。\",\n \"selectModelLabel\": \"默认分析模型\",\n \"selectModelPlaceholder\": \"请选择分析模型\",\n \"selectModelFirst\": \"请先选择分析模型\",\n \"noModelOptions\": \"当前没有可用的文本模型，请先在个人设置中配置 Provider 和模型。\",\n \"saveFailed\": \"保存默认分析模型失败，请稍后重试\"\n },\n \"sidebar\": {\n \"dragToMove\": \"拖动调整位置\",\n \"listTitle\": \"剧集列表\",\n \"episodeCount\": \"{count}集\",\n \"empty\": \"暂无剧集，点击下方创建\",\n \"save\": \"保存\",\n \"deleteConfirm\": \"确定删除「{name}」？\",\n \"delete\": \"删除\",\n \"cancel\": \"取消\",\n \"rename\": \"重命名\",\n \"newEpisodePlaceholder\": \"输入剧集名称...\",\n \"create\": \"创建\",\n \"addEpisode\": \"添加剧集\"\n }\n}\n"
},
{
"path": "messages/zh/worldContextModal.json",
"content": "{\n \"title\": \"世界观与人设\",\n \"description\": \"定义全局通用的角色外貌、场景风格和环境描述\",\n \"placeholder\": \"例如：\\n【男主】张三，25岁，黑色短发，总是穿着一件洗得发白的牛仔夹克，眼神忧郁。\\n【女主】李四，22岁，红发双马尾，性格活泼,喜欢穿洛丽塔风格的裙子。\\n【场景】2077年的赛博朋克城市，霓虹灯闪烁，终年下雨...\",\n \"hint\": \"这些设定将被所有剧集继承，作为AI绘画的基础参考。\"\n}"
},
{
"path": "middleware.ts",
"content": "import createMiddleware from 'next-intl/middleware';\nimport { routing } from './src/i18n/routing';\n\nexport default createMiddleware(routing);\n\nexport const config = {\n // 匹配所有路径，除了 api、_next/static、_next/image、favicon.ico 等\n matcher: [\n // 匹配所有路径\n '/((?!api|m|_next/static|_next/image|favicon.ico|.*\\\\.png|.*\\\\.jpg|.*\\\\.jpeg|.*\\\\.svg|.*\\\\.gif|.*\\\\.ico).*)'\n ]\n};\n"
},
{
"path": "next.config.ts",
"content": "import type { NextConfig } from \"next\";\nimport createNextIntlPlugin from 'next-intl/plugin';\n\nconst withNextIntl = createNextIntlPlugin('./src/i18n.ts');\n\nconst nextConfig: NextConfig = {\n // 已删除 ignoreBuildErrors / ignoreDuringBuilds，构建保持严格门禁\n // Next 15 的 allowedDevOrigins 是顶层配置，不属于 experimental\n allowedDevOrigins: [\n 'http://192.168.31.218:3000',\n 'http://192.168.31.*:3000',\n ],\n};\n\nexport default withNextIntl(nextConfig);\n"
},
{
"path": "package.json",
"content": "{\n \"name\": \"waoowaoo\",\n \"version\": \"0.3.0\",\n \"private\": true,\n \"engines\": {\n \"node\": \">=18.18.0\",\n \"npm\": \">=9.0.0\"\n },\n \"scripts\": {\n \"postinstall\": \"prisma generate\",\n \"prepare\": \"husky\",\n \"dev\": \"npm run storage:init && concurrently --kill-others \\\"npm run dev:next\\\" \\\"npm run dev:worker\\\" \\\"npm run dev:watchdog\\\" \\\"npm run dev:board\\\"\",\n \"dev:next\": \"cross-env NODE_OPTIONS=\\\"--no-deprecation\\\" next dev --turbopack -H 0.0.0.0\",\n \"dev:worker\": \"tsx watch --env-file=.env src/lib/workers/index.ts\",\n \"dev:watchdog\": \"tsx watch --env-file=.env scripts/watchdog.ts\",\n \"dev:board\": \"tsx watch --env-file=.env scripts/bull-board.ts\",\n \"dev:turbo\": \"next dev --turbopack -H 0.0.0.0\",\n \"build\": \"prisma generate && next build\",\n \"build:turbo\": \"next build --turbopack\",\n \"start\": \"npm run storage:init && concurrently --kill-others \\\"npm run start:next\\\" \\\"npm run start:worker\\\" \\\"npm run start:watchdog\\\" \\\"npm run start:board\\\"\",\n \"storage:init\": \"tsx --env-file=.env src/lib/storage/init.ts\",\n \"start:next\": \"next start -H 0.0.0.0\",\n \"start:worker\": \"tsx --env-file=.env src/lib/workers/index.ts\",\n \"start:watchdog\": \"tsx --env-file=.env scripts/watchdog.ts\",\n \"start:board\": \"tsx --env-file=.env scripts/bull-board.ts\",\n \"stats:errors\": \"tsx scripts/task-error-stats.ts\",\n \"check:api-handler\": \"node scripts/guards/api-route-contract-guard.mjs\",\n \"check:logs\": \"tsx scripts/check-no-console.ts\",\n \"check:log-semantic\": \"tsx scripts/check-log-semantic.ts\",\n \"check:media-normalization\": \"tsx scripts/check-media-normalization.ts\",\n \"check:no-api-direct-llm-call\": \"node scripts/guards/no-api-direct-llm-call.mjs\",\n \"check:no-internal-task-sync-fallback\": \"node scripts/guards/no-internal-task-sync-fallback.mjs\",\n \"check:no-media-provider-bypass\": \"node scripts/guards/no-media-provider-bypass.mjs\",\n \"check:no-model-key-downgrade\": \"node scripts/guards/no-model-key-downgrade.mjs\",\n \"check:no-provider-guessing\": \"node scripts/guards/no-provider-guessing.mjs\",\n \"check:no-hardcoded-model-capabilities\": \"node scripts/guards/no-hardcoded-model-capabilities.mjs\",\n \"check:capability-catalog\": \"node scripts/check-capability-catalog.mjs\",\n \"check:pricing-catalog\": \"node scripts/check-pricing-catalog.mjs\",\n \"check:model-config-contract\": \"node scripts/check-model-config-contract.mjs --strict\",\n \"check:config-center-guards\": \"npm run check:no-model-key-downgrade && npm run check:no-provider-guessing && npm run check:no-hardcoded-model-capabilities && npm run check:capability-catalog && npm run check:pricing-catalog\",\n \"check:outbound-image-unification\": \"tsx scripts/check-outbound-image-unification.ts\",\n \"check:outbound-image-runtime-sample\": \"tsx scripts/check-outbound-image-runtime-sample.ts\",\n \"check:outbound-image-success-rate\": \"tsx scripts/check-outbound-image-success-rate.ts\",\n \"check:image-urls-contract\": \"tsx scripts/check-image-urls-contract.ts\",\n \"verify:outbound-image\": \"cross-env BILLING_TEST_BOOTSTRAP=0 vitest run src/lib/media/outbound-image.test.ts src/lib/media/image-url.test.ts && npm run check:outbound-image-unification\",\n \"check:task-loading\": \"node scripts/guards/task-loading-guard.mjs\",\n \"check:task-target-states-no-polling\": \"node scripts/guards/task-target-states-no-polling-guard.mjs\",\n \"check:no-server-mirror-state\": \"node scripts/guards/no-server-mirror-state.mjs\",\n \"check:no-multiple-sources-of-truth\": \"node scripts/guards/no-multiple-sources-of-truth.mjs\",\n \"check:file-line-count\": \"node scripts/guards/file-line-count-guard.mjs\",\n \"check:no-duplicate-endpoint-entry\": \"node scripts/guards/no-duplicate-endpoint-entry.mjs\",\n \"check:test-route-coverage\": \"node scripts/guards/test-route-coverage-guard.mjs\",\n \"check:test-tasktype-coverage\": \"node scripts/guards/test-tasktype-coverage-guard.mjs\",\n \"check:test-behavior-quality\": \"node scripts/guards/test-behavior-quality-guard.mjs\",\n \"check:changed-test-impact\": \"node scripts/guards/changed-file-test-impact-guard.mjs\",\n \"check:test-behavior-route-coverage\": \"node scripts/guards/test-behavior-route-coverage-guard.mjs\",\n \"check:test-behavior-tasktype-coverage\": \"node scripts/guards/test-behavior-tasktype-coverage-guard.mjs\",\n \"check:test-coverage-guards\": \"npm run check:test-behavior-quality && npm run check:test-route-coverage && npm run check:test-tasktype-coverage && npm run check:test-behavior-route-coverage && npm run check:test-behavior-tasktype-coverage\",\n \"check:requirements-matrix\": \"cross-env BILLING_TEST_BOOTSTRAP=0 vitest run tests/contracts/requirements-matrix.test.ts\",\n \"check:image-reference-normalization\": \"node scripts/guards/image-reference-normalization-guard.mjs\",\n \"check:task-submit-compensation\": \"node scripts/guards/task-submit-compensation-guard.mjs\",\n \"check:locale-navigation\": \"node scripts/guards/locale-navigation-guard.mjs\",\n \"check:prompt-i18n\": \"node scripts/guards/prompt-i18n-guard.mjs\",\n \"check:prompt-i18n-regression\": \"node scripts/guards/prompt-semantic-regression.mjs\",\n \"check:prompt-ab-regression\": \"node scripts/guards/prompt-ab-regression.mjs\",\n \"check:prompt-json-canary\": \"node scripts/guards/prompt-json-canary-guard.mjs\",\n \"billing:cleanup-pending-freezes\": \"tsx scripts/billing-cleanup-pending-freezes.ts\",\n \"billing:reconcile-ledger\": \"tsx scripts/billing-reconcile-ledger.ts\",\n \"cleanup:remove-legacy-voice-data\": \"tsx scripts/cleanup-remove-legacy-voice-data.ts\",\n \"test:billing\": \"npm run test:billing:coverage\",\n \"test:billing:unit\": \"cross-env BILLING_TEST_BOOTSTRAP=0 vitest run tests/unit/billing\",\n \"test:billing:integration\": \"cross-env BILLING_TEST_BOOTSTRAP=1 vitest run tests/integration/billing\",\n \"test:billing:concurrency\": \"cross-env BILLING_TEST_BOOTSTRAP=1 vitest run tests/concurrency/billing\",\n \"test:billing:coverage\": \"cross-env BILLING_TEST_BOOTSTRAP=1 vitest run --coverage tests/unit/billing tests/integration/billing tests/concurrency/billing\",\n \"test:guards\": \"npm run check:api-handler && npm run check:image-reference-normalization && npm run check:task-submit-compensation && npm run check:no-api-direct-llm-call && npm run check:test-coverage-guards && npm run check:changed-test-impact && npm run check:requirements-matrix && npm run check:locale-navigation\",\n \"test:unit:all\": \"cross-env BILLING_TEST_BOOTSTRAP=0 vitest run tests/unit\",\n \"test:integration:api\": \"cross-env BILLING_TEST_BOOTSTRAP=1 vitest run tests/integration/api\",\n \"test:integration:provider\": \"cross-env BILLING_TEST_BOOTSTRAP=0 vitest run tests/integration/provider\",\n \"test:integration:chain\": \"cross-env BILLING_TEST_BOOTSTRAP=1 vitest run tests/integration/chain\",\n \"test:integration:task\": \"cross-env BILLING_TEST_BOOTSTRAP=1 vitest run tests/integration/task\",\n \"test:system\": \"cross-env SYSTEM_TEST_BOOTSTRAP=1 vitest run tests/system\",\n \"test:regression:cases\": \"cross-env SYSTEM_TEST_BOOTSTRAP=1 vitest run tests/regression\",\n \"test:behavior:unit\": \"cross-env BILLING_TEST_BOOTSTRAP=0 vitest run tests/unit/helpers tests/unit/worker tests/unit/optimistic tests/unit/guards\",\n \"test:behavior:api\": \"cross-env BILLING_TEST_BOOTSTRAP=0 vitest run tests/integration/api/contract\",\n \"test:behavior:provider\": \"cross-env BILLING_TEST_BOOTSTRAP=0 vitest run tests/integration/provider\",\n \"test:behavior:chain\": \"cross-env BILLING_TEST_BOOTSTRAP=0 vitest run tests/integration/chain\",\n \"test:behavior:guards\": \"npm run check:api-handler && npm run check:image-reference-normalization && npm run check:task-submit-compensation && npm run check:test-coverage-guards && npm run check:requirements-matrix\",\n \"test:behavior:full\": \"npm run test:behavior:guards && npm run test:behavior:unit && npm run test:behavior:api && npm run test:behavior:provider && npm run test:behavior:chain\",\n \"test:coverage:core-baseline\": \"cross-env SYSTEM_TEST_BOOTSTRAP=1 vitest run --config vitest.core-coverage.config.ts tests/unit tests/integration/api tests/integration/provider tests/integration/chain tests/integration/task tests/system tests/regression\",\n \"test:all\": \"npm run test:guards && npm run test:unit:all && npm run test:billing:integration && npm run test:billing:concurrency && npm run test:integration:api && npm run test:integration:provider && npm run test:integration:chain && npm run test:integration:task && npm run test:system && npm run test:regression:cases\",\n \"test:regression\": \"npm run test:all\",\n \"test:pr\": \"bash scripts/test-regression-runner.sh npm run test:all\",\n \"typecheck\": \"tsc --noEmit\",\n \"lint:all\": \"npm run lint -- .\",\n \"verify:commit\": \"npm run lint:all && npm run typecheck && npm run test:all\",\n \"verify:push\": \"npm run lint:all && npm run typecheck && npm run test:all && npm run build\",\n \"migrate:image-urls-contract\": \"tsx scripts/migrate-image-urls-contract.ts\",\n \"migrate:model-config-contract\": \"tsx scripts/migrations/migrate-model-config-contract.ts\",\n \"migrate:capability-selections\": \"tsx scripts/migrations/migrate-capability-selections.ts\",\n \"migrate:gateway-route-openai-compat\": \"tsx scripts/migrations/migrate-gateway-route-openai-compat.ts\",\n \"migrate:custom-pricing-v2\": \"tsx scripts/migrations/migrate-custom-pricing-v2.ts\",\n \"migrate:graph-artifacts-unique-index\": \"tsx scripts/migrations/migrate-graph-artifacts-unique-index.ts\",\n \"migrate:release-blockers\": \"tsx scripts/migrations/migrate-release-blockers.ts\",\n \"backup:media-safety\": \"tsx scripts/media-safety-backup.ts\",\n \"backup:media-restore-dry-run\": \"tsx scripts/media-restore-dry-run.ts\",\n \"media:backfill-refs\": \"tsx scripts/media-backfill-refs.ts\",\n \"backup:archive-legacy-media\": \"tsx scripts/media-archive-legacy-refs.ts\",\n \"backup:unreferenced-media-index\": \"tsx scripts/media-build-unreferenced-index.ts\",\n \"lint\": \"eslint\",\n \"clean\": \"rimraf .next\"\n },\n \"dependencies\": {\n \"@ai-sdk/google\": \"^3.0.22\",\n \"@ai-sdk/openai\": \"^3.0.26\",\n \"@ai-sdk/react\": \"^3.0.118\",\n \"@aws-sdk/client-s3\": \"^3.883.0\",\n \"@aws-sdk/s3-request-presigner\": \"^3.883.0\",\n \"@bull-board/api\": \"^6.16.4\",\n \"@bull-board/express\": \"^6.16.4\",\n \"@dnd-kit/core\": \"^6.3.1\",\n \"@dnd-kit/sortable\": \"^10.0.0\",\n \"@fal-ai/client\": \"^1.7.2\",\n \"@google/genai\": \"^1.34.0\",\n \"@next-auth/prisma-adapter\": \"^1.0.7\",\n \"@openrouter/sdk\": \"^0.3.11\",\n \"@prisma/client\": \"^6.19.2\",\n \"@remotion/cli\": \"^4.0.405\",\n \"@remotion/player\": \"^4.0.405\",\n \"@tanstack/react-query\": \"^5.90.20\",\n \"@types/archiver\": \"^7.0.0\",\n \"@types/bcryptjs\": \"^3.0.0\",\n \"@types/express\": \"^5.0.6\",\n \"@vercel/og\": \"^0.8.6\",\n \"ai\": \"^6.0.116\",\n \"archiver\": \"^7.0.1\",\n \"bcryptjs\": \"^3.0.2\",\n \"bullmq\": \"^5.67.3\",\n \"cos-nodejs-sdk-v5\": \"^2.15.4\",\n \"express\": \"^5.2.1\",\n \"file-saver\": \"^2.0.5\",\n \"ioredis\": \"^5.9.2\",\n \"jsonrepair\": \"^3.13.2\",\n \"jszip\": \"^3.10.1\",\n \"lru-cache\": \"^11.2.6\",\n \"lucide-react\": \"^0.575.0\",\n \"mammoth\": \"^1.11.0\",\n \"mysql2\": \"^3.15.1\",\n \"next\": \"^15.5.7\",\n \"next-auth\": \"^4.24.11\",\n \"next-intl\": \"^4.7.0\",\n \"openai\": \"^6.8.1\",\n \"prisma\": \"^6.19.2\",\n \"react\": \"^19.1.2\",\n \"react-dom\": \"^19.1.2\",\n \"react-grab\": \"^0.1.20\",\n \"react-hot-toast\": \"^2.6.0\",\n \"remotion\": \"^4.0.405\",\n \"sharp\": \"^0.34.5\",\n \"undici\": \"^7.22.0\",\n \"zod\": \"^3.25.76\"\n },\n \"devDependencies\": {\n \"@eslint/eslintrc\": \"^3\",\n \"@tailwindcss/postcss\": \"^4\",\n \"@types/file-saver\": \"^2.0.7\",\n \"@types/node\": \"^20\",\n \"@types/react\": \"^19\",\n \"@types/react-dom\": \"^19\",\n \"@vitest/coverage-v8\": \"^2.1.8\",\n \"concurrently\": \"^9.2.1\",\n \"cross-env\": \"^10.1.0\",\n \"eslint\": \"^9\",\n \"eslint-config-next\": \"15.5.4\",\n \"husky\": \"^9.1.7\",\n \"rimraf\": \"^6.1.2\",\n \"tailwindcss\": \"^4\",\n \"tsx\": \"^4.20.5\",\n \"typescript\": \"^5\",\n \"vitest\": \"^2.1.8\"\n }\n}\n"
},
{
"path": "postcss.config.mjs",
"content": "const config = {\n plugins: [\"@tailwindcss/postcss\"],\n};\n\nexport default config;\n"
},
{
"path": "prisma/schema.prisma",
"content": "generator client {\n provider = \"prisma-client-js\"\n}\n\ndatasource db {\n provider = \"mysql\"\n url = env(\"DATABASE_URL\")\n}\n\nmodel Account {\n id String @id @default(uuid())\n userId String\n type String\n provider String\n providerAccountId String\n refresh_token String? @db.Text\n access_token String? @db.Text\n expires_at Int?\n token_type String?\n scope String?\n id_token String? @db.Text\n session_state String?\n user User @relation(fields: [userId], references: [id], onDelete: Cascade)\n\n @@unique([provider, providerAccountId])\n @@index([userId])\n @@map(\"account\")\n}\n\nmodel CharacterAppearance {\n id String @id @default(uuid())\n characterId String\n appearanceIndex Int\n changeReason String\n description String? @db.Text\n descriptions String? @db.Text\n imageUrl String? @db.Text\n imageUrls String? @db.Text\n selectedIndex Int?\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n previousImageUrl String? @db.Text\n previousImageUrls String? @db.Text\n previousDescription String? @db.Text // 上一次的描述词（用于撤回）\n previousDescriptions String? @db.Text // 上一次的描述词数组（用于撤回）\n imageMediaId String?\n imageMedia MediaObject? @relation(\"CharacterAppearanceImageMedia\", fields: [imageMediaId], references: [id], onDelete: SetNull)\n character NovelPromotionCharacter @relation(fields: [characterId], references: [id], onDelete: Cascade)\n\n @@unique([characterId, appearanceIndex])\n @@index([characterId])\n @@index([imageMediaId])\n @@map(\"character_appearances\")\n}\n\nmodel LocationImage {\n id String @id @default(uuid())\n locationId String\n imageIndex Int\n description String? @db.Text\n imageUrl String? @db.Text\n isSelected Boolean @default(false)\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n previousImageUrl String? @db.Text\n previousDescription String? @db.Text // 上一次的描述词（用于撤回）\n imageMediaId String?\n imageMedia MediaObject? @relation(\"LocationImageMedia\", fields: [imageMediaId], references: [id], onDelete: SetNull)\n location NovelPromotionLocation @relation(\"LocationImages\", fields: [locationId], references: [id], onDelete: Cascade)\n selectedByLocations NovelPromotionLocation[] @relation(\"SelectedLocationImage\")\n\n @@unique([locationId, imageIndex])\n @@index([locationId])\n @@index([imageMediaId])\n @@map(\"location_images\")\n}\n\nmodel NovelPromotionCharacter {\n id String @id @default(uuid())\n novelPromotionProjectId String\n name String\n aliases String? @db.Text\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n customVoiceUrl String? @db.Text\n customVoiceMediaId String?\n customVoiceMedia MediaObject? @relation(\"NovelPromotionCharacterVoiceMedia\", fields: [customVoiceMediaId], references: [id], onDelete: SetNull)\n voiceId String?\n voiceType String?\n profileData String? @db.Text\n profileConfirmed Boolean @default(false)\n introduction String? @db.Text // 角色介绍（身份、关系、称呼映射，如\"我\"对应此角色）\n sourceGlobalCharacterId String? // 🆕 来源全局角色ID（复制时记录）\n appearances CharacterAppearance[]\n novelPromotionProject NovelPromotionProject @relation(fields: [novelPromotionProjectId], references: [id], onDelete: Cascade)\n\n @@index([novelPromotionProjectId])\n @@index([customVoiceMediaId])\n @@map(\"novel_promotion_characters\")\n}\n\nmodel NovelPromotionLocation {\n id String @id @default(uuid())\n novelPromotionProjectId String\n name String\n summary String? @db.Text // 场景简要描述（用途/人物关联）\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n sourceGlobalLocationId String? // 🆕 来源全局场景ID（复制时记录）\n selectedImageId String?\n selectedImage LocationImage? @relation(\"SelectedLocationImage\", fields: [selectedImageId], references: [id], onDelete: SetNull)\n images LocationImage[] @relation(\"LocationImages\")\n novelPromotionProject NovelPromotionProject @relation(fields: [novelPromotionProjectId], references: [id], onDelete: Cascade)\n\n @@index([novelPromotionProjectId])\n @@map(\"novel_promotion_locations\")\n}\n\nmodel NovelPromotionEpisode {\n id String @id @default(uuid())\n novelPromotionProjectId String\n episodeNumber Int\n name String\n description String? @db.Text\n novelText String? @db.Text\n audioUrl String? @db.Text\n audioMediaId String?\n audioMedia MediaObject? @relation(\"NovelPromotionEpisodeAudioMedia\", fields: [audioMediaId], references: [id], onDelete: SetNull)\n srtContent String? @db.Text\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n speakerVoices String? @db.Text\n clips NovelPromotionClip[]\n novelPromotionProject NovelPromotionProject @relation(fields: [novelPromotionProjectId], references: [id], onDelete: Cascade)\n shots NovelPromotionShot[]\n storyboards NovelPromotionStoryboard[]\n voiceLines NovelPromotionVoiceLine[]\n editorProject VideoEditorProject?\n\n @@unique([novelPromotionProjectId, episodeNumber])\n @@index([novelPromotionProjectId])\n @@index([audioMediaId])\n @@map(\"novel_promotion_episodes\")\n}\n\n// 视频编辑器项目 - 存储剪辑数据\nmodel VideoEditorProject {\n id String @id @default(uuid())\n episodeId String @unique\n projectData String @db.Text // JSON 存储编辑项目数据\n renderStatus String? // pending | rendering | completed | failed\n renderTaskId String?\n outputUrl String? @db.Text\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n episode NovelPromotionEpisode @relation(fields: [episodeId], references: [id], onDelete: Cascade)\n\n @@map(\"video_editor_projects\")\n}\n\nmodel NovelPromotionClip {\n id String @id @default(uuid())\n episodeId String\n start Int?\n end Int?\n duration Int?\n summary String @db.Text\n location String? @db.Text\n content String @db.Text\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n characters String? @db.Text\n endText String? @db.Text\n shotCount Int?\n startText String? @db.Text\n screenplay String? @db.Text\n episode NovelPromotionEpisode @relation(fields: [episodeId], references: [id], onDelete: Cascade)\n shots NovelPromotionShot[]\n storyboard NovelPromotionStoryboard?\n\n @@index([episodeId])\n @@map(\"novel_promotion_clips\")\n}\n\nmodel NovelPromotionPanel {\n id String @id @default(uuid())\n storyboardId String\n panelIndex Int\n panelNumber Int?\n shotType String? @db.Text\n cameraMove String? @db.Text\n description String? @db.Text\n location String? @db.Text\n characters String? @db.Text\n srtSegment String? @db.Text\n srtStart Float?\n srtEnd Float?\n duration Float?\n imagePrompt String? @db.Text\n imageUrl String? @db.Text\n imageMediaId String?\n imageMedia MediaObject? @relation(\"NovelPromotionPanelImageMedia\", fields: [imageMediaId], references: [id], onDelete: SetNull)\n imageHistory String? @db.Text\n videoPrompt String? @db.Text\n firstLastFramePrompt String? @db.Text\n videoUrl String? @db.Text\n videoGenerationMode String? @db.Text // 视频生成方式：normal | firstlastframe\n videoMediaId String?\n videoMedia MediaObject? @relation(\"NovelPromotionPanelVideoMedia\", fields: [videoMediaId], references: [id], onDelete: SetNull)\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n sceneType String?\n candidateImages String? @db.Text\n linkedToNextPanel Boolean @default(false)\n lipSyncTaskId String?\n lipSyncVideoUrl String?\n lipSyncVideoMediaId String?\n lipSyncVideoMedia MediaObject? @relation(\"NovelPromotionPanelLipSyncVideoMedia\", fields: [lipSyncVideoMediaId], references: [id], onDelete: SetNull)\n sketchImageUrl String? @db.Text\n sketchImageMediaId String?\n sketchImageMedia MediaObject? @relation(\"NovelPromotionPanelSketchMedia\", fields: [sketchImageMediaId], references: [id], onDelete: SetNull)\n photographyRules String? @db.Text\n actingNotes String? @db.Text // 演技指导数据 JSON\n previousImageUrl String? @db.Text\n previousImageMediaId String?\n previousImageMedia MediaObject? @relation(\"NovelPromotionPanelPreviousImageMedia\", fields: [previousImageMediaId], references: [id], onDelete: SetNull)\n storyboard NovelPromotionStoryboard @relation(fields: [storyboardId], references: [id], onDelete: Cascade)\n matchedVoiceLines NovelPromotionVoiceLine[]\n\n @@unique([storyboardId, panelIndex])\n @@index([storyboardId])\n @@index([imageMediaId])\n @@index([videoMediaId])\n @@index([lipSyncVideoMediaId])\n @@index([sketchImageMediaId])\n @@index([previousImageMediaId])\n @@map(\"novel_promotion_panels\")\n}\n\nmodel NovelPromotionProject {\n id String @id @default(uuid())\n projectId String @unique\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n analysisModel String? // 用户配置的分析模型（nullable，必须配置后才能使用）\n imageModel String? // 用户配置的图片模型\n videoModel String? // 用户配置的视频模型\n audioModel String? // 用户配置的语音模型\n videoRatio String @default(\"9:16\")\n ttsRate String @default(\"+50%\")\n globalAssetText String? @db.Text\n artStyle String @default(\"american-comic\")\n artStylePrompt String? @db.Text\n characterModel String? // 用户配置的角色图片模型\n locationModel String? // 用户配置的场景图片模型\n storyboardModel String? // 用户配置的分镜图片模型\n editModel String? // 用户配置的修图/编辑模型\n videoResolution String @default(\"720p\")\n capabilityOverrides String? @db.Text\n workflowMode String @default(\"srt\")\n lastEpisodeId String?\n imageResolution String @default(\"2K\")\n importStatus String?\n characters NovelPromotionCharacter[]\n episodes NovelPromotionEpisode[]\n locations NovelPromotionLocation[]\n project Project @relation(fields: [projectId], references: [id], onDelete: Cascade)\n\n @@map(\"novel_promotion_projects\")\n}\n\nmodel NovelPromotionShot {\n id String @id @default(uuid())\n episodeId String\n clipId String?\n shotId String\n srtStart Int\n srtEnd Int\n srtDuration Float\n sequence String? @db.Text\n locations String? @db.Text\n characters String? @db.Text\n plot String? @db.Text\n imagePrompt String? @db.Text\n scale String? @db.Text\n module String? @db.Text\n focus String? @db.Text\n zhSummarize String? @db.Text\n imageUrl String? @db.Text\n imageMediaId String?\n imageMedia MediaObject? @relation(\"NovelPromotionShotImageMedia\", fields: [imageMediaId], references: [id], onDelete: SetNull)\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n pov String? @db.Text\n clip NovelPromotionClip? @relation(fields: [clipId], references: [id], onDelete: Cascade)\n episode NovelPromotionEpisode @relation(fields: [episodeId], references: [id], onDelete: Cascade)\n\n @@index([clipId])\n @@index([episodeId])\n @@index([shotId])\n @@index([imageMediaId])\n @@map(\"novel_promotion_shots\")\n}\n\nmodel NovelPromotionStoryboard {\n id String @id @default(uuid())\n episodeId String\n clipId String @unique\n storyboardImageUrl String? @db.Text\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n panelCount Int @default(9)\n storyboardTextJson String? @db.Text\n imageHistory String? @db.Text\n candidateImages String? @db.Text\n lastError String?\n photographyPlan String? @db.Text\n panels NovelPromotionPanel[]\n clip NovelPromotionClip @relation(fields: [clipId], references: [id], onDelete: Cascade)\n episode NovelPromotionEpisode @relation(fields: [episodeId], references: [id], onDelete: Cascade)\n supplementaryPanels SupplementaryPanel[]\n\n @@index([clipId])\n @@index([episodeId])\n @@map(\"novel_promotion_storyboards\")\n}\n\nmodel SupplementaryPanel {\n id String @id @default(uuid())\n storyboardId String\n sourceType String\n sourcePanelId String?\n description String? @db.Text\n imagePrompt String? @db.Text\n imageUrl String? @db.Text\n imageMediaId String?\n imageMedia MediaObject? @relation(\"SupplementaryPanelImageMedia\", fields: [imageMediaId], references: [id], onDelete: SetNull)\n characters String? @db.Text\n location String? @db.Text\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n storyboard NovelPromotionStoryboard @relation(fields: [storyboardId], references: [id], onDelete: Cascade)\n\n @@index([storyboardId])\n @@index([imageMediaId])\n @@map(\"supplementary_panels\")\n}\n\nmodel Project {\n id String @id @default(uuid())\n name String\n description String? @db.Text\n mode String @default(\"novel-promotion\")\n userId String\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n lastAccessedAt DateTime?\n novelPromotionData NovelPromotionProject?\n user User @relation(fields: [userId], references: [id], onDelete: Cascade)\n usageCosts UsageCost[]\n\n @@index([userId])\n @@map(\"projects\")\n}\n\nmodel Session {\n id String @id @default(uuid())\n sessionToken String @unique(map: \"Session_sessionToken_key\")\n userId String\n expires DateTime\n user User @relation(fields: [userId], references: [id], onDelete: Cascade)\n\n @@index([userId])\n @@map(\"session\")\n}\n\nmodel UsageCost {\n id String @id @default(uuid())\n projectId String\n userId String\n apiType String\n model String\n action String\n quantity Int\n unit String\n cost Decimal @db.Decimal(18, 6)\n metadata String? @db.Text\n createdAt DateTime @default(now())\n project Project @relation(fields: [projectId], references: [id], onDelete: Cascade)\n user User @relation(fields: [userId], references: [id], onDelete: Cascade)\n\n @@index([apiType])\n @@index([createdAt])\n @@index([projectId])\n @@index([userId])\n @@map(\"usage_costs\")\n}\n\nmodel User {\n id String @id @default(uuid())\n name String @unique(map: \"User_name_key\")\n email String?\n emailVerified DateTime?\n image String?\n password String?\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n accounts Account[]\n projects Project[]\n sessions Session[]\n usageCosts UsageCost[]\n balance UserBalance?\n preferences UserPreference?\n\n // 资产中心\n globalAssetFolders GlobalAssetFolder[]\n globalCharacters GlobalCharacter[]\n globalLocations GlobalLocation[]\n globalVoices GlobalVoice[]\n tasks Task[]\n taskEvents TaskEvent[]\n graphRuns GraphRun[]\n graphEvents GraphEvent[]\n\n @@map(\"user\")\n}\n\nmodel UserPreference {\n id String @id @default(uuid())\n userId String @unique\n analysisModel String? // 用户配置的分析模型（nullable，必须配置后才能使用）\n characterModel String? // 用户配置的角色图片模型\n locationModel String? // 用户配置的场景图片模型\n storyboardModel String? // 用户配置的分镜图片模型\n editModel String? // 用户配置的修图模型\n videoModel String? // 用户配置的视频模型\n audioModel String? // 用户配置的语音模型\n lipSyncModel String? // 用户配置的口型同步模型\n voiceDesignModel String? // 用户配置的音色设计模型\n analysisConcurrency Int? // 分析流程并发上限\n imageConcurrency Int? // 图像流程并发上限\n videoConcurrency Int? // 视频流程并发上限\n videoRatio String @default(\"9:16\")\n videoResolution String @default(\"720p\")\n artStyle String @default(\"american-comic\")\n ttsRate String @default(\"+50%\")\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n imageResolution String @default(\"2K\")\n capabilityDefaults String? @db.Text\n\n // API Key 配置（极简版）\n llmBaseUrl String? @default(\"https://openrouter.ai/api/v1\")\n llmApiKey String? @db.Text // 加密存储\n falApiKey String? @db.Text // FAL（图片+视频+语音）\n googleAiKey String? @db.Text // Google AI（Gemini 图片）\n arkApiKey String? @db.Text // 火山引擎（Seedream+Seedance）\n qwenApiKey String? @db.Text // 阿里百炼（声音设计）\n\n // 自定义模型列表 + 价格（JSON）\n customModels String? @db.Text\n\n // 自定义 OpenAI 兼容提供商列表（JSON，包含加密的 API Key）\n customProviders String? @db.Text\n\n user User @relation(fields: [userId], references: [id], onDelete: Cascade)\n\n @@map(\"user_preferences\")\n}\n\nmodel VerificationToken {\n identifier String\n token String @unique(map: \"VerificationToken_token_key\")\n expires DateTime\n\n @@unique([identifier, token])\n @@map(\"verificationtoken\")\n}\n\nmodel NovelPromotionVoiceLine {\n id String @id @default(uuid())\n episodeId String\n lineIndex Int\n speaker String\n content String @db.Text\n voicePresetId String?\n audioUrl String? @db.Text\n audioMediaId String?\n audioMedia MediaObject? @relation(\"NovelPromotionVoiceLineAudioMedia\", fields: [audioMediaId], references: [id], onDelete: SetNull)\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n emotionPrompt String? @db.Text\n emotionStrength Float? @default(0.4)\n matchedPanelIndex Int?\n matchedStoryboardId String?\n audioDuration Int?\n matchedPanelId String?\n episode NovelPromotionEpisode @relation(fields: [episodeId], references: [id], onDelete: Cascade)\n matchedPanel NovelPromotionPanel? @relation(fields: [matchedPanelId], references: [id])\n\n @@unique([episodeId, lineIndex])\n @@index([episodeId])\n @@index([matchedPanelId])\n @@index([audioMediaId])\n @@map(\"novel_promotion_voice_lines\")\n}\n\nmodel VoicePreset {\n id String @id @default(uuid())\n name String\n audioUrl String @db.Text\n audioMediaId String?\n audioMedia MediaObject? @relation(\"VoicePresetAudioMedia\", fields: [audioMediaId], references: [id], onDelete: SetNull)\n description String? @db.Text\n gender String?\n isSystem Boolean @default(true)\n createdAt DateTime @default(now())\n\n @@index([audioMediaId])\n @@map(\"voice_presets\")\n}\n\nmodel UserBalance {\n id String @id @default(uuid())\n userId String @unique\n balance Decimal @default(0) @db.Decimal(18, 6)\n frozenAmount Decimal @default(0) @db.Decimal(18, 6)\n totalSpent Decimal @default(0) @db.Decimal(18, 6)\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n user User @relation(fields: [userId], references: [id], onDelete: Cascade)\n\n @@map(\"user_balances\")\n}\n\nmodel BalanceFreeze {\n id String @id @default(uuid())\n userId String\n amount Decimal @db.Decimal(18, 6)\n status String @default(\"pending\")\n source String? @db.VarChar(64)\n taskId String?\n requestId String?\n idempotencyKey String? @unique\n metadata String? @db.Text\n expiresAt DateTime?\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n\n @@index([userId])\n @@index([status])\n @@index([taskId])\n @@map(\"balance_freezes\")\n}\n\nmodel BalanceTransaction {\n id String @id @default(uuid())\n userId String\n type String\n amount Decimal @db.Decimal(18, 6)\n balanceAfter Decimal @db.Decimal(18, 6)\n description String? @db.Text\n relatedId String?\n freezeId String?\n operatorId String? @db.VarChar(64)\n externalOrderId String? @db.VarChar(128)\n idempotencyKey String? @db.VarChar(128)\n projectId String? @db.VarChar(128) // 关联项目 ID，用于流水展示项目名\n episodeId String? @db.VarChar(128) // 关联集数 ID，用于流水展示集数\n taskType String? @db.VarChar(64) // 任务类型 key（与 action 一致），用于前端 i18n\n billingMeta String? @db.Text // 计费详情 JSON: { quantity, unit, model, resolution, duration, tokens... }\n createdAt DateTime @default(now())\n\n @@index([userId])\n @@index([type])\n @@index([createdAt])\n @@index([freezeId])\n @@index([externalOrderId])\n @@index([projectId])\n @@unique([userId, type, idempotencyKey])\n @@map(\"balance_transactions\")\n}\n\nmodel Task {\n id String @id @default(uuid())\n userId String\n projectId String\n episodeId String?\n type String\n targetType String\n targetId String\n status String @default(\"queued\")\n progress Int @default(0)\n attempt Int @default(0)\n maxAttempts Int @default(5)\n priority Int @default(0)\n dedupeKey String? @unique\n externalId String?\n payload Json?\n result Json?\n errorCode String?\n errorMessage String? @db.Text\n billingInfo Json?\n billedAt DateTime?\n queuedAt DateTime @default(now())\n startedAt DateTime?\n finishedAt DateTime?\n heartbeatAt DateTime?\n enqueuedAt DateTime?\n enqueueAttempts Int @default(0)\n lastEnqueueError String? @db.Text\n createdAt DateTime @default(now())\n updatedAt DateTime @updatedAt\n\n user User @relation(fields: [userId], references: [id], onDelete: Cascade)\n events TaskEvent[]\n\n @@index([status])\n @@index([type])\n @@index([targetType, targetId])\n @@index([projectId])\n @@index([userId])\n @@index([heartbeatAt])\n @@map(\"tasks\")\n}\n\nmodel TaskEvent {\n id Int @id @default(autoincrement())\n taskId String\n projectId String\n userId String\n eventType String\n payload Json?\n createdAt DateTime @default(now())\n\n task Task @relation(fields: [taskId], references: [id], onDelete: Cascade)\n user User @relation(fields: [userId], references: [id], onDelete: Cascade)\n\n @@index([projectId, id])\n @@index([taskId])\n @@index([userId])\n @@map(\"task_events\")\n}\n\nmodel GraphRun {\n id String @id @default(uuid())\n userId String\n projectId String\n episodeId String?\n workflowType String\n taskType String?\n taskId String? @unique\n targetType String\n targetId String\n status String @default(\"queued\")\n input Json?\n output Json?\n errorCode String?\n errorMessage String? @db.Text\n cancelRequestedAt DateTime?\n leaseOwner String?\n leaseExpiresAt DateTime?\n heartbeatAt DateTime?\n workflowVersion Int @default(1)\n queuedAt DateTime @default(now())\n startedAt DateTime?\n finishedAt DateTime?\n lastSeq Int @default(0)\n createdAt DateTime @default(now())\n updatedAt DateTime @updatedAt\n user User @relation(fields: [userId], references: [id], onDelete: Cascade)\n steps GraphStep[]\n attempts GraphStepAttempt[]\n events GraphEvent[]\n checkpoints GraphCheckpoint[]\n artifacts GraphArtifact[]\n\n @@index([projectId, status])\n @@index([userId, createdAt])\n @@index([taskId])\n @@index([targetType, targetId])\n @@index([workflowType, targetType, targetId, status])\n @@index([leaseExpiresAt])\n @@map(\"graph_runs\")\n}\n\nmodel GraphStep {\n id String @id @default(uuid())\n runId String\n stepKey String\n stepTitle String\n status String @default(\"pending\")\n currentAttempt Int @default(0)\n stepIndex Int\n stepTotal Int\n startedAt DateTime?\n finishedAt DateTime?\n lastErrorCode String?\n lastErrorMessage String? @db.Text\n createdAt DateTime @default(now())\n updatedAt DateTime @updatedAt\n run GraphRun @relation(fields: [runId], references: [id], onDelete: Cascade)\n attempts GraphStepAttempt[]\n\n @@unique([runId, stepKey])\n @@index([runId, status])\n @@index([runId, stepIndex])\n @@map(\"graph_steps\")\n}\n\nmodel GraphStepAttempt {\n id String @id @default(uuid())\n runId String\n stepKey String\n attempt Int\n status String @default(\"pending\")\n provider String?\n modelKey String?\n inputHash String?\n input Json?\n outputText String? @db.Text\n outputReasoning String? @db.Text\n usageJson Json?\n errorCode String?\n errorMessage String? @db.Text\n startedAt DateTime?\n finishedAt DateTime?\n createdAt DateTime @default(now())\n updatedAt DateTime @updatedAt\n run GraphRun @relation(fields: [runId], references: [id], onDelete: Cascade)\n step GraphStep @relation(fields: [runId, stepKey], references: [runId, stepKey], onDelete: Cascade)\n\n @@unique([runId, stepKey, attempt])\n @@index([runId, stepKey])\n @@index([runId, createdAt])\n @@map(\"graph_step_attempts\")\n}\n\nmodel GraphEvent {\n id BigInt @id @default(autoincrement())\n runId String\n projectId String\n userId String\n seq Int\n eventType String\n stepKey String?\n attempt Int?\n lane String?\n payload Json?\n createdAt DateTime @default(now())\n run GraphRun @relation(fields: [runId], references: [id], onDelete: Cascade)\n user User @relation(fields: [userId], references: [id], onDelete: Cascade)\n\n @@unique([runId, seq])\n @@index([projectId, id])\n @@index([runId, id])\n @@index([userId, id])\n @@map(\"graph_events\")\n}\n\nmodel GraphCheckpoint {\n id String @id @default(uuid())\n runId String\n nodeKey String\n version Int\n stateJson Json\n stateBytes Int\n createdAt DateTime @default(now())\n run GraphRun @relation(fields: [runId], references: [id], onDelete: Cascade)\n\n @@unique([runId, nodeKey, version])\n @@index([runId, createdAt])\n @@map(\"graph_checkpoints\")\n}\n\nmodel GraphArtifact {\n id String @id @default(uuid())\n runId String\n stepKey String?\n artifactType String\n refId String\n versionHash String?\n payload Json?\n createdAt DateTime @default(now())\n run GraphRun @relation(fields: [runId], references: [id], onDelete: Cascade)\n\n @@unique([runId, stepKey, artifactType, refId])\n @@index([runId])\n @@index([runId, stepKey])\n @@index([artifactType, refId])\n @@map(\"graph_artifacts\")\n}\n\n// ==================== 资产中心 ====================\n\n// 资产文件夹（一层，不支持嵌套）\nmodel GlobalAssetFolder {\n id String @id @default(uuid())\n userId String\n name String\n createdAt DateTime @default(now())\n updatedAt DateTime @updatedAt\n\n user User @relation(fields: [userId], references: [id], onDelete: Cascade)\n characters GlobalCharacter[]\n locations GlobalLocation[]\n voices GlobalVoice[]\n\n @@index([userId])\n @@map(\"global_asset_folders\")\n}\n\n// 全局角色（结构与 NovelPromotionCharacter 一致）\nmodel GlobalCharacter {\n id String @id @default(uuid())\n userId String\n folderId String?\n name String\n aliases String? @db.Text\n profileData String? @db.Text\n profileConfirmed Boolean @default(false)\n voiceId String?\n voiceType String?\n customVoiceUrl String? @db.Text\n customVoiceMediaId String?\n customVoiceMedia MediaObject? @relation(\"GlobalCharacterVoiceMedia\", fields: [customVoiceMediaId], references: [id], onDelete: SetNull)\n globalVoiceId String? // 绑定的全局音色 ID\n createdAt DateTime @default(now())\n updatedAt DateTime @updatedAt\n\n user User @relation(fields: [userId], references: [id], onDelete: Cascade)\n folder GlobalAssetFolder? @relation(fields: [folderId], references: [id], onDelete: SetNull)\n appearances GlobalCharacterAppearance[]\n\n @@index([userId])\n @@index([folderId])\n @@index([customVoiceMediaId])\n @@map(\"global_characters\")\n}\n\n// 全局角色形象（结构与 CharacterAppearance 一致）\nmodel GlobalCharacterAppearance {\n id String @id @default(uuid())\n characterId String\n appearanceIndex Int\n changeReason String @default(\"default\")\n artStyle String?\n description String? @db.Text\n descriptions String? @db.Text\n imageUrl String? @db.Text\n imageMediaId String?\n imageMedia MediaObject? @relation(\"GlobalCharacterAppearanceImageMedia\", fields: [imageMediaId], references: [id], onDelete: SetNull)\n imageUrls String? @db.Text\n selectedIndex Int?\n previousImageUrl String? @db.Text\n previousImageMediaId String?\n previousImageMedia MediaObject? @relation(\"GlobalCharacterAppearancePreviousImageMedia\", fields: [previousImageMediaId], references: [id], onDelete: SetNull)\n previousImageUrls String? @db.Text\n previousDescription String? @db.Text // 上一次的描述词（用于撤回）\n previousDescriptions String? @db.Text // 上一次的描述词数组（用于撤回）\n createdAt DateTime @default(now())\n updatedAt DateTime @updatedAt\n\n character GlobalCharacter @relation(fields: [characterId], references: [id], onDelete: Cascade)\n\n @@unique([characterId, appearanceIndex])\n @@index([characterId])\n @@index([imageMediaId])\n @@index([previousImageMediaId])\n @@map(\"global_character_appearances\")\n}\n\n// 全局场景（结构与 NovelPromotionLocation 一致）\nmodel GlobalLocation {\n id String @id @default(uuid())\n userId String\n folderId String?\n name String\n artStyle String?\n summary String? @db.Text\n createdAt DateTime @default(now())\n updatedAt DateTime @updatedAt\n\n user User @relation(fields: [userId], references: [id], onDelete: Cascade)\n folder GlobalAssetFolder? @relation(fields: [folderId], references: [id], onDelete: SetNull)\n images GlobalLocationImage[]\n\n @@index([userId])\n @@index([folderId])\n @@map(\"global_locations\")\n}\n\n// 全局场景图片（结构与 LocationImage 一致）\nmodel GlobalLocationImage {\n id String @id @default(uuid())\n locationId String\n imageIndex Int\n description String? @db.Text\n imageUrl String? @db.Text\n imageMediaId String?\n imageMedia MediaObject? @relation(\"GlobalLocationImageMedia\", fields: [imageMediaId], references: [id], onDelete: SetNull)\n isSelected Boolean @default(false)\n previousImageUrl String? @db.Text\n previousImageMediaId String?\n previousImageMedia MediaObject? @relation(\"GlobalLocationImagePreviousImageMedia\", fields: [previousImageMediaId], references: [id], onDelete: SetNull)\n previousDescription String? @db.Text // 上一次的描述词（用于撤回）\n createdAt DateTime @default(now())\n updatedAt DateTime @updatedAt\n\n location GlobalLocation @relation(fields: [locationId], references: [id], onDelete: Cascade)\n\n @@unique([locationId, imageIndex])\n @@index([locationId])\n @@index([imageMediaId])\n @@index([previousImageMediaId])\n @@map(\"global_location_images\")\n}\n\n// 全局音色库\nmodel GlobalVoice {\n id String @id @default(uuid())\n userId String\n folderId String?\n name String // 音色名称\n description String? @db.Text // 详细描述\n voiceId String? // qwen-tts-vd 的 voice ID\n voiceType String @default(\"qwen-designed\") // qwen-designed | custom\n customVoiceUrl String? @db.Text // 上传的音频 URL（预览用）\n customVoiceMediaId String?\n customVoiceMedia MediaObject? @relation(\"GlobalVoiceCustomVoiceMedia\", fields: [customVoiceMediaId], references: [id], onDelete: SetNull)\n voicePrompt String? @db.Text // AI 设计时的提示词\n gender String? // male | female | neutral\n language String @default(\"zh\")\n createdAt DateTime @default(now())\n updatedAt DateTime @updatedAt\n\n user User @relation(fields: [userId], references: [id], onDelete: Cascade)\n folder GlobalAssetFolder? @relation(fields: [folderId], references: [id], onDelete: SetNull)\n\n @@index([userId])\n @@index([folderId])\n @@index([customVoiceMediaId])\n @@map(\"global_voices\")\n}\n\nmodel MediaObject {\n id String @id @default(uuid())\n publicId String @unique\n storageKey String @unique @db.VarChar(512)\n sha256 String?\n mimeType String?\n sizeBytes BigInt?\n width Int?\n height Int?\n durationMs Int?\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n\n characterAppearanceImages CharacterAppearance[] @relation(\"CharacterAppearanceImageMedia\")\n locationImages LocationImage[] @relation(\"LocationImageMedia\")\n novelPromotionCharacterVoices NovelPromotionCharacter[] @relation(\"NovelPromotionCharacterVoiceMedia\")\n novelPromotionEpisodeAudios NovelPromotionEpisode[] @relation(\"NovelPromotionEpisodeAudioMedia\")\n novelPromotionPanelImages NovelPromotionPanel[] @relation(\"NovelPromotionPanelImageMedia\")\n novelPromotionPanelVideos NovelPromotionPanel[] @relation(\"NovelPromotionPanelVideoMedia\")\n novelPromotionPanelLipSyncVideos NovelPromotionPanel[] @relation(\"NovelPromotionPanelLipSyncVideoMedia\")\n novelPromotionPanelSketchImages NovelPromotionPanel[] @relation(\"NovelPromotionPanelSketchMedia\")\n novelPromotionPanelPreviousImages NovelPromotionPanel[] @relation(\"NovelPromotionPanelPreviousImageMedia\")\n novelPromotionShotImages NovelPromotionShot[] @relation(\"NovelPromotionShotImageMedia\")\n supplementaryPanelImages SupplementaryPanel[] @relation(\"SupplementaryPanelImageMedia\")\n novelPromotionVoiceLineAudios NovelPromotionVoiceLine[] @relation(\"NovelPromotionVoiceLineAudioMedia\")\n voicePresetAudios VoicePreset[] @relation(\"VoicePresetAudioMedia\")\n globalCharacterVoices GlobalCharacter[] @relation(\"GlobalCharacterVoiceMedia\")\n globalCharacterAppearanceImages GlobalCharacterAppearance[] @relation(\"GlobalCharacterAppearanceImageMedia\")\n globalCharacterAppearancePreviousImgs GlobalCharacterAppearance[] @relation(\"GlobalCharacterAppearancePreviousImageMedia\")\n globalLocationImageImages GlobalLocationImage[] @relation(\"GlobalLocationImageMedia\")\n globalLocationImagePreviousImages GlobalLocationImage[] @relation(\"GlobalLocationImagePreviousImageMedia\")\n globalVoiceCustomVoices GlobalVoice[] @relation(\"GlobalVoiceCustomVoiceMedia\")\n\n @@index([createdAt])\n @@map(\"media_objects\")\n}\n\nmodel LegacyMediaRefBackup {\n id String @id @default(uuid())\n runId String\n tableName String\n rowId String\n fieldName String\n legacyValue String @db.Text\n checksum String\n createdAt DateTime @default(now())\n\n @@index([runId])\n @@index([tableName, fieldName])\n @@map(\"legacy_media_refs_backup\")\n}\n"
},
{
"path": "prisma/schema.sqlit.prisma",
"content": "generator client {\n provider = \"prisma-client-js\"\n}\n\ndatasource db {\n provider = \"sqlite\"\n url = env(\"DATABASE_URL\")\n}\n\nmodel Account {\n id String @id @default(uuid())\n userId String\n type String\n provider String\n providerAccountId String\n refresh_token String?\n access_token String?\n expires_at Int?\n token_type String?\n scope String?\n id_token String?\n session_state String?\n user User @relation(fields: [userId], references: [id], onDelete: Cascade)\n\n @@unique([provider, providerAccountId])\n @@index([userId])\n @@map(\"account\")\n}\n\nmodel CharacterAppearance {\n id String @id @default(uuid())\n characterId String\n appearanceIndex Int\n changeReason String\n description String?\n descriptions String?\n imageUrl String?\n imageUrls String?\n selectedIndex Int?\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n previousImageUrl String?\n previousImageUrls String?\n previousDescription String? // 上一次的描述词（用于撤回）\n previousDescriptions String? // 上一次的描述词数组（用于撤回）\n imageMediaId String?\n imageMedia MediaObject? @relation(\"CharacterAppearanceImageMedia\", fields: [imageMediaId], references: [id], onDelete: SetNull)\n character NovelPromotionCharacter @relation(fields: [characterId], references: [id], onDelete: Cascade)\n\n @@unique([characterId, appearanceIndex])\n @@index([characterId])\n @@index([imageMediaId])\n @@map(\"character_appearances\")\n}\n\nmodel LocationImage {\n id String @id @default(uuid())\n locationId String\n imageIndex Int\n description String?\n imageUrl String?\n isSelected Boolean @default(false)\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n previousImageUrl String?\n previousDescription String? // 上一次的描述词（用于撤回）\n imageMediaId String?\n imageMedia MediaObject? @relation(\"LocationImageMedia\", fields: [imageMediaId], references: [id], onDelete: SetNull)\n location NovelPromotionLocation @relation(fields: [locationId], references: [id], onDelete: Cascade)\n\n @@unique([locationId, imageIndex])\n @@index([locationId])\n @@index([imageMediaId])\n @@map(\"location_images\")\n}\n\nmodel NovelPromotionCharacter {\n id String @id @default(uuid())\n novelPromotionProjectId String\n name String\n aliases String?\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n customVoiceUrl String?\n customVoiceMediaId String?\n customVoiceMedia MediaObject? @relation(\"NovelPromotionCharacterVoiceMedia\", fields: [customVoiceMediaId], references: [id], onDelete: SetNull)\n voiceId String?\n voiceType String?\n profileData String?\n profileConfirmed Boolean @default(false)\n introduction String? // 角色介绍（身份、关系、称呼映射，如\"我\"对应此角色）\n sourceGlobalCharacterId String? // 🆕 来源全局角色ID（复制时记录）\n appearances CharacterAppearance[]\n novelPromotionProject NovelPromotionProject @relation(fields: [novelPromotionProjectId], references: [id], onDelete: Cascade)\n\n @@index([novelPromotionProjectId])\n @@index([customVoiceMediaId])\n @@map(\"novel_promotion_characters\")\n}\n\nmodel NovelPromotionLocation {\n id String @id @default(uuid())\n novelPromotionProjectId String\n name String\n summary String? // 场景简要描述（用途/人物关联）\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n sourceGlobalLocationId String? // 🆕 来源全局场景ID（复制时记录）\n images LocationImage[]\n novelPromotionProject NovelPromotionProject @relation(fields: [novelPromotionProjectId], references: [id], onDelete: Cascade)\n\n @@index([novelPromotionProjectId])\n @@map(\"novel_promotion_locations\")\n}\n\nmodel NovelPromotionEpisode {\n id String @id @default(uuid())\n novelPromotionProjectId String\n episodeNumber Int\n name String\n description String?\n novelText String?\n audioUrl String?\n audioMediaId String?\n audioMedia MediaObject? @relation(\"NovelPromotionEpisodeAudioMedia\", fields: [audioMediaId], references: [id], onDelete: SetNull)\n srtContent String?\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n speakerVoices String?\n clips NovelPromotionClip[]\n novelPromotionProject NovelPromotionProject @relation(fields: [novelPromotionProjectId], references: [id], onDelete: Cascade)\n shots NovelPromotionShot[]\n storyboards NovelPromotionStoryboard[]\n voiceLines NovelPromotionVoiceLine[]\n editorProject VideoEditorProject?\n\n @@unique([novelPromotionProjectId, episodeNumber])\n @@index([novelPromotionProjectId])\n @@index([audioMediaId])\n @@map(\"novel_promotion_episodes\")\n}\n\n// 视频编辑器项目 - 存储剪辑数据\nmodel VideoEditorProject {\n id String @id @default(uuid())\n episodeId String @unique\n projectData String // JSON 存储编辑项目数据\n renderStatus String? // pending | rendering | completed | failed\n renderTaskId String?\n outputUrl String?\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n episode NovelPromotionEpisode @relation(fields: [episodeId], references: [id], onDelete: Cascade)\n\n @@map(\"video_editor_projects\")\n}\n\nmodel NovelPromotionClip {\n id String @id @default(uuid())\n episodeId String\n start Int?\n end Int?\n duration Int?\n summary String\n location String?\n content String\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n characters String?\n endText String?\n shotCount Int?\n startText String?\n screenplay String?\n episode NovelPromotionEpisode @relation(fields: [episodeId], references: [id], onDelete: Cascade)\n shots NovelPromotionShot[]\n storyboard NovelPromotionStoryboard?\n\n @@index([episodeId])\n @@map(\"novel_promotion_clips\")\n}\n\nmodel NovelPromotionPanel {\n id String @id @default(uuid())\n storyboardId String\n panelIndex Int\n panelNumber Int?\n shotType String?\n cameraMove String?\n description String?\n location String?\n characters String?\n srtSegment String?\n srtStart Float?\n srtEnd Float?\n duration Float?\n imagePrompt String?\n imageUrl String?\n imageMediaId String?\n imageMedia MediaObject? @relation(\"NovelPromotionPanelImageMedia\", fields: [imageMediaId], references: [id], onDelete: SetNull)\n imageHistory String?\n videoPrompt String?\n firstLastFramePrompt String?\n videoUrl String?\n videoGenerationMode String? // 视频生成方式：normal | firstlastframe\n videoMediaId String?\n videoMedia MediaObject? @relation(\"NovelPromotionPanelVideoMedia\", fields: [videoMediaId], references: [id], onDelete: SetNull)\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n sceneType String?\n candidateImages String?\n linkedToNextPanel Boolean @default(false)\n lipSyncTaskId String?\n lipSyncVideoUrl String?\n lipSyncVideoMediaId String?\n lipSyncVideoMedia MediaObject? @relation(\"NovelPromotionPanelLipSyncVideoMedia\", fields: [lipSyncVideoMediaId], references: [id], onDelete: SetNull)\n sketchImageUrl String?\n sketchImageMediaId String?\n sketchImageMedia MediaObject? @relation(\"NovelPromotionPanelSketchMedia\", fields: [sketchImageMediaId], references: [id], onDelete: SetNull)\n photographyRules String?\n actingNotes String? // 演技指导数据 JSON\n previousImageUrl String?\n previousImageMediaId String?\n previousImageMedia MediaObject? @relation(\"NovelPromotionPanelPreviousImageMedia\", fields: [previousImageMediaId], references: [id], onDelete: SetNull)\n storyboard NovelPromotionStoryboard @relation(fields: [storyboardId], references: [id], onDelete: Cascade)\n matchedVoiceLines NovelPromotionVoiceLine[]\n\n @@unique([storyboardId, panelIndex])\n @@index([storyboardId])\n @@index([imageMediaId])\n @@index([videoMediaId])\n @@index([lipSyncVideoMediaId])\n @@index([sketchImageMediaId])\n @@index([previousImageMediaId])\n @@map(\"novel_promotion_panels\")\n}\n\nmodel NovelPromotionProject {\n id String @id @default(uuid())\n projectId String @unique\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n analysisModel String? // 用户配置的分析模型（nullable，必须配置后才能使用）\n imageModel String? // 用户配置的图片模型\n videoModel String? // 用户配置的视频模型\n audioModel String? // 用户配置的语音模型\n videoRatio String @default(\"9:16\")\n ttsRate String @default(\"+50%\")\n globalAssetText String?\n artStyle String @default(\"american-comic\")\n artStylePrompt String?\n characterModel String? // 用户配置的角色图片模型\n locationModel String? // 用户配置的场景图片模型\n storyboardModel String? // 用户配置的分镜图片模型\n editModel String? // 用户配置的修图/编辑模型\n videoResolution String @default(\"720p\")\n capabilityOverrides String?\n workflowMode String @default(\"srt\")\n lastEpisodeId String?\n imageResolution String @default(\"2K\")\n importStatus String?\n characters NovelPromotionCharacter[]\n episodes NovelPromotionEpisode[]\n locations NovelPromotionLocation[]\n project Project @relation(fields: [projectId], references: [id], onDelete: Cascade)\n\n @@map(\"novel_promotion_projects\")\n}\n\nmodel NovelPromotionShot {\n id String @id @default(uuid())\n episodeId String\n clipId String?\n shotId String\n srtStart Int\n srtEnd Int\n srtDuration Float\n sequence String?\n locations String?\n characters String?\n plot String?\n imagePrompt String?\n scale String?\n module String?\n focus String?\n zhSummarize String?\n imageUrl String?\n imageMediaId String?\n imageMedia MediaObject? @relation(\"NovelPromotionShotImageMedia\", fields: [imageMediaId], references: [id], onDelete: SetNull)\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n pov String?\n clip NovelPromotionClip? @relation(fields: [clipId], references: [id], onDelete: Cascade)\n episode NovelPromotionEpisode @relation(fields: [episodeId], references: [id], onDelete: Cascade)\n\n @@index([clipId])\n @@index([episodeId])\n @@index([shotId])\n @@index([imageMediaId])\n @@map(\"novel_promotion_shots\")\n}\n\nmodel NovelPromotionStoryboard {\n id String @id @default(uuid())\n episodeId String\n clipId String @unique\n storyboardImageUrl String?\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n panelCount Int @default(9)\n storyboardTextJson String?\n imageHistory String?\n candidateImages String?\n lastError String?\n photographyPlan String?\n panels NovelPromotionPanel[]\n clip NovelPromotionClip @relation(fields: [clipId], references: [id], onDelete: Cascade)\n episode NovelPromotionEpisode @relation(fields: [episodeId], references: [id], onDelete: Cascade)\n supplementaryPanels SupplementaryPanel[]\n\n @@index([clipId])\n @@index([episodeId])\n @@map(\"novel_promotion_storyboards\")\n}\n\nmodel SupplementaryPanel {\n id String @id @default(uuid())\n storyboardId String\n sourceType String\n sourcePanelId String?\n description String?\n imagePrompt String?\n imageUrl String?\n imageMediaId String?\n imageMedia MediaObject? @relation(\"SupplementaryPanelImageMedia\", fields: [imageMediaId], references: [id], onDelete: SetNull)\n characters String?\n location String?\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n storyboard NovelPromotionStoryboard @relation(fields: [storyboardId], references: [id], onDelete: Cascade)\n\n @@index([storyboardId])\n @@index([imageMediaId])\n @@map(\"supplementary_panels\")\n}\n\nmodel Project {\n id String @id @default(uuid())\n name String\n description String?\n mode String @default(\"novel-promotion\")\n userId String\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n lastAccessedAt DateTime?\n novelPromotionData NovelPromotionProject?\n user User @relation(fields: [userId], references: [id], onDelete: Cascade)\n usageCosts UsageCost[]\n\n @@index([userId])\n @@map(\"projects\")\n}\n\nmodel Session {\n id String @id @default(uuid())\n sessionToken String @unique(map: \"Session_sessionToken_key\")\n userId String\n expires DateTime\n user User @relation(fields: [userId], references: [id], onDelete: Cascade)\n\n @@index([userId])\n @@map(\"session\")\n}\n\nmodel UsageCost {\n id String @id @default(uuid())\n projectId String\n userId String\n apiType String\n model String\n action String\n quantity Int\n unit String\n cost Decimal\n metadata String?\n createdAt DateTime @default(now())\n project Project @relation(fields: [projectId], references: [id], onDelete: Cascade)\n user User @relation(fields: [userId], references: [id], onDelete: Cascade)\n\n @@index([apiType])\n @@index([createdAt])\n @@index([projectId])\n @@index([userId])\n @@map(\"usage_costs\")\n}\n\nmodel User {\n id String @id @default(uuid())\n name String @unique(map: \"User_name_key\")\n email String?\n emailVerified DateTime?\n image String?\n password String?\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n accounts Account[]\n projects Project[]\n sessions Session[]\n usageCosts UsageCost[]\n balance UserBalance?\n preferences UserPreference?\n\n // 资产中心\n globalAssetFolders GlobalAssetFolder[]\n globalCharacters GlobalCharacter[]\n globalLocations GlobalLocation[]\n globalVoices GlobalVoice[]\n tasks Task[]\n taskEvents TaskEvent[]\n\n @@map(\"user\")\n}\n\nmodel UserPreference {\n id String @id @default(uuid())\n userId String @unique\n analysisModel String? // 用户配置的分析模型（nullable，必须配置后才能使用）\n characterModel String? // 用户配置的角色图片模型\n locationModel String? // 用户配置的场景图片模型\n storyboardModel String? // 用户配置的分镜图片模型\n editModel String? // 用户配置的修图模型\n videoModel String? // 用户配置的视频模型\n audioModel String? // 用户配置的语音模型\n lipSyncModel String? // 用户配置的口型同步模型\n voiceDesignModel String? // 用户配置的音色设计模型\n analysisConcurrency Int? // 分析流程并发上限\n imageConcurrency Int? // 图像流程并发上限\n videoConcurrency Int? // 视频流程并发上限\n videoRatio String @default(\"9:16\")\n videoResolution String @default(\"720p\")\n artStyle String @default(\"american-comic\")\n ttsRate String @default(\"+50%\")\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n imageResolution String @default(\"2K\")\n capabilityDefaults String?\n\n // API Key 配置（极简版）\n llmBaseUrl String? @default(\"https://openrouter.ai/api/v1\")\n llmApiKey String? // 加密存储\n falApiKey String? // FAL（图片+视频+语音）\n googleAiKey String? // Google AI（Gemini 图片）\n arkApiKey String? // 火山引擎（Seedream+Seedance）\n qwenApiKey String? // 阿里百炼（声音设计）\n\n // 自定义模型列表 + 价格（JSON）\n customModels String?\n\n // 自定义 OpenAI 兼容提供商列表（JSON，包含加密的 API Key）\n customProviders String?\n\n user User @relation(fields: [userId], references: [id], onDelete: Cascade)\n\n @@map(\"user_preferences\")\n}\n\nmodel VerificationToken {\n identifier String\n token String @unique(map: \"VerificationToken_token_key\")\n expires DateTime\n\n @@unique([identifier, token])\n @@map(\"verificationtoken\")\n}\n\nmodel NovelPromotionVoiceLine {\n id String @id @default(uuid())\n episodeId String\n lineIndex Int\n speaker String\n content String\n voicePresetId String?\n audioUrl String?\n audioMediaId String?\n audioMedia MediaObject? @relation(\"NovelPromotionVoiceLineAudioMedia\", fields: [audioMediaId], references: [id], onDelete: SetNull)\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n emotionPrompt String?\n emotionStrength Float? @default(0.4)\n matchedPanelIndex Int?\n matchedStoryboardId String?\n audioDuration Int?\n matchedPanelId String?\n episode NovelPromotionEpisode @relation(fields: [episodeId], references: [id], onDelete: Cascade)\n matchedPanel NovelPromotionPanel? @relation(fields: [matchedPanelId], references: [id])\n\n @@unique([episodeId, lineIndex])\n @@index([episodeId])\n @@index([matchedPanelId])\n @@index([audioMediaId])\n @@map(\"novel_promotion_voice_lines\")\n}\n\nmodel VoicePreset {\n id String @id @default(uuid())\n name String\n audioUrl String\n audioMediaId String?\n audioMedia MediaObject? @relation(\"VoicePresetAudioMedia\", fields: [audioMediaId], references: [id], onDelete: SetNull)\n description String?\n gender String?\n isSystem Boolean @default(true)\n createdAt DateTime @default(now())\n\n @@index([audioMediaId])\n @@map(\"voice_presets\")\n}\n\nmodel UserBalance {\n id String @id @default(uuid())\n userId String @unique\n balance Decimal @default(0)\n frozenAmount Decimal @default(0)\n totalSpent Decimal @default(0)\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n user User @relation(fields: [userId], references: [id], onDelete: Cascade)\n\n @@map(\"user_balances\")\n}\n\nmodel BalanceFreeze {\n id String @id @default(uuid())\n userId String\n amount Decimal\n status String @default(\"pending\")\n source String?\n taskId String?\n requestId String?\n idempotencyKey String? @unique\n metadata String?\n expiresAt DateTime?\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n\n @@index([userId])\n @@index([status])\n @@index([taskId])\n @@map(\"balance_freezes\")\n}\n\nmodel BalanceTransaction {\n id String @id @default(uuid())\n userId String\n type String\n amount Decimal\n balanceAfter Decimal\n description String?\n relatedId String?\n freezeId String?\n operatorId String?\n externalOrderId String?\n idempotencyKey String?\n createdAt DateTime @default(now())\n\n @@index([userId])\n @@index([type])\n @@index([createdAt])\n @@index([freezeId])\n @@index([externalOrderId])\n @@unique([userId, type, idempotencyKey])\n @@map(\"balance_transactions\")\n}\n\nmodel Task {\n id String @id @default(uuid())\n userId String\n projectId String\n episodeId String?\n type String\n targetType String\n targetId String\n status String @default(\"queued\")\n progress Int @default(0)\n attempt Int @default(0)\n maxAttempts Int @default(5)\n priority Int @default(0)\n dedupeKey String? @unique\n externalId String?\n payload Json?\n result Json?\n errorCode String?\n errorMessage String?\n billingInfo Json?\n billedAt DateTime?\n queuedAt DateTime @default(now())\n startedAt DateTime?\n finishedAt DateTime?\n heartbeatAt DateTime?\n enqueuedAt DateTime?\n enqueueAttempts Int @default(0)\n lastEnqueueError String?\n createdAt DateTime @default(now())\n updatedAt DateTime @updatedAt\n\n user User @relation(fields: [userId], references: [id], onDelete: Cascade)\n events TaskEvent[]\n\n @@index([status])\n @@index([type])\n @@index([targetType, targetId])\n @@index([projectId])\n @@index([userId])\n @@index([heartbeatAt])\n @@map(\"tasks\")\n}\n\nmodel TaskEvent {\n id Int @id @default(autoincrement())\n taskId String\n projectId String\n userId String\n eventType String\n payload Json?\n createdAt DateTime @default(now())\n\n task Task @relation(fields: [taskId], references: [id], onDelete: Cascade)\n user User @relation(fields: [userId], references: [id], onDelete: Cascade)\n\n @@index([projectId, id])\n @@index([taskId])\n @@index([userId])\n @@map(\"task_events\")\n}\n\n// ==================== 资产中心 ====================\n\n// 资产文件夹（一层，不支持嵌套）\nmodel GlobalAssetFolder {\n id String @id @default(uuid())\n userId String\n name String\n createdAt DateTime @default(now())\n updatedAt DateTime @updatedAt\n\n user User @relation(fields: [userId], references: [id], onDelete: Cascade)\n characters GlobalCharacter[]\n locations GlobalLocation[]\n voices GlobalVoice[]\n\n @@index([userId])\n @@map(\"global_asset_folders\")\n}\n\n// 全局角色（结构与 NovelPromotionCharacter 一致）\nmodel GlobalCharacter {\n id String @id @default(uuid())\n userId String\n folderId String?\n name String\n aliases String?\n profileData String?\n profileConfirmed Boolean @default(false)\n voiceId String?\n voiceType String?\n customVoiceUrl String?\n customVoiceMediaId String?\n customVoiceMedia MediaObject? @relation(\"GlobalCharacterVoiceMedia\", fields: [customVoiceMediaId], references: [id], onDelete: SetNull)\n globalVoiceId String? // 绑定的全局音色 ID\n createdAt DateTime @default(now())\n updatedAt DateTime @updatedAt\n\n user User @relation(fields: [userId], references: [id], onDelete: Cascade)\n folder GlobalAssetFolder? @relation(fields: [folderId], references: [id], onDelete: SetNull)\n appearances GlobalCharacterAppearance[]\n\n @@index([userId])\n @@index([folderId])\n @@index([customVoiceMediaId])\n @@map(\"global_characters\")\n}\n\n// 全局角色形象（结构与 CharacterAppearance 一致）\nmodel GlobalCharacterAppearance {\n id String @id @default(uuid())\n characterId String\n appearanceIndex Int\n changeReason String @default(\"default\")\n artStyle String?\n description String?\n descriptions String?\n imageUrl String?\n imageMediaId String?\n imageMedia MediaObject? @relation(\"GlobalCharacterAppearanceImageMedia\", fields: [imageMediaId], references: [id], onDelete: SetNull)\n imageUrls String?\n selectedIndex Int?\n previousImageUrl String?\n previousImageMediaId String?\n previousImageMedia MediaObject? @relation(\"GlobalCharacterAppearancePreviousImageMedia\", fields: [previousImageMediaId], references: [id], onDelete: SetNull)\n previousImageUrls String?\n previousDescription String? // 上一次的描述词（用于撤回）\n previousDescriptions String? // 上一次的描述词数组（用于撤回）\n createdAt DateTime @default(now())\n updatedAt DateTime @updatedAt\n\n character GlobalCharacter @relation(fields: [characterId], references: [id], onDelete: Cascade)\n\n @@unique([characterId, appearanceIndex])\n @@index([characterId])\n @@index([imageMediaId])\n @@index([previousImageMediaId])\n @@map(\"global_character_appearances\")\n}\n\n// 全局场景（结构与 NovelPromotionLocation 一致）\nmodel GlobalLocation {\n id String @id @default(uuid())\n userId String\n folderId String?\n name String\n artStyle String?\n summary String?\n createdAt DateTime @default(now())\n updatedAt DateTime @updatedAt\n\n user User @relation(fields: [userId], references: [id], onDelete: Cascade)\n folder GlobalAssetFolder? @relation(fields: [folderId], references: [id], onDelete: SetNull)\n images GlobalLocationImage[]\n\n @@index([userId])\n @@index([folderId])\n @@map(\"global_locations\")\n}\n\n// 全局场景图片（结构与 LocationImage 一致）\nmodel GlobalLocationImage {\n id String @id @default(uuid())\n locationId String\n imageIndex Int\n description String?\n imageUrl String?\n imageMediaId String?\n imageMedia MediaObject? @relation(\"GlobalLocationImageMedia\", fields: [imageMediaId], references: [id], onDelete: SetNull)\n isSelected Boolean @default(false)\n previousImageUrl String?\n previousImageMediaId String?\n previousImageMedia MediaObject? @relation(\"GlobalLocationImagePreviousImageMedia\", fields: [previousImageMediaId], references: [id], onDelete: SetNull)\n previousDescription String? // 上一次的描述词（用于撤回）\n createdAt DateTime @default(now())\n updatedAt DateTime @updatedAt\n\n location GlobalLocation @relation(fields: [locationId], references: [id], onDelete: Cascade)\n\n @@unique([locationId, imageIndex])\n @@index([locationId])\n @@index([imageMediaId])\n @@index([previousImageMediaId])\n @@map(\"global_location_images\")\n}\n\n// 全局音色库\nmodel GlobalVoice {\n id String @id @default(uuid())\n userId String\n folderId String?\n name String // 音色名称\n description String? // 详细描述\n voiceId String? // qwen-tts-vd 的 voice ID\n voiceType String @default(\"qwen-designed\") // qwen-designed | custom\n customVoiceUrl String? // 上传的音频 URL（预览用）\n customVoiceMediaId String?\n customVoiceMedia MediaObject? @relation(\"GlobalVoiceCustomVoiceMedia\", fields: [customVoiceMediaId], references: [id], onDelete: SetNull)\n voicePrompt String? // AI 设计时的提示词\n gender String? // male | female | neutral\n language String @default(\"zh\")\n createdAt DateTime @default(now())\n updatedAt DateTime @updatedAt\n\n user User @relation(fields: [userId], references: [id], onDelete: Cascade)\n folder GlobalAssetFolder? @relation(fields: [folderId], references: [id], onDelete: SetNull)\n\n @@index([userId])\n @@index([folderId])\n @@index([customVoiceMediaId])\n @@map(\"global_voices\")\n}\n\nmodel MediaObject {\n id String @id @default(uuid())\n publicId String @unique\n storageKey String @unique\n sha256 String?\n mimeType String?\n sizeBytes BigInt?\n width Int?\n height Int?\n durationMs Int?\n createdAt DateTime @default(now())\n updatedAt DateTime @default(now()) @updatedAt\n\n characterAppearanceImages CharacterAppearance[] @relation(\"CharacterAppearanceImageMedia\")\n locationImages LocationImage[] @relation(\"LocationImageMedia\")\n novelPromotionCharacterVoices NovelPromotionCharacter[] @relation(\"NovelPromotionCharacterVoiceMedia\")\n novelPromotionEpisodeAudios NovelPromotionEpisode[] @relation(\"NovelPromotionEpisodeAudioMedia\")\n novelPromotionPanelImages NovelPromotionPanel[] @relation(\"NovelPromotionPanelImageMedia\")\n novelPromotionPanelVideos NovelPromotionPanel[] @relation(\"NovelPromotionPanelVideoMedia\")\n novelPromotionPanelLipSyncVideos NovelPromotionPanel[] @relation(\"NovelPromotionPanelLipSyncVideoMedia\")\n novelPromotionPanelSketchImages NovelPromotionPanel[] @relation(\"NovelPromotionPanelSketchMedia\")\n novelPromotionPanelPreviousImages NovelPromotionPanel[] @relation(\"NovelPromotionPanelPreviousImageMedia\")\n novelPromotionShotImages NovelPromotionShot[] @relation(\"NovelPromotionShotImageMedia\")\n supplementaryPanelImages SupplementaryPanel[] @relation(\"SupplementaryPanelImageMedia\")\n novelPromotionVoiceLineAudios NovelPromotionVoiceLine[] @relation(\"NovelPromotionVoiceLineAudioMedia\")\n voicePresetAudios VoicePreset[] @relation(\"VoicePresetAudioMedia\")\n globalCharacterVoices GlobalCharacter[] @relation(\"GlobalCharacterVoiceMedia\")\n globalCharacterAppearanceImages GlobalCharacterAppearance[] @relation(\"GlobalCharacterAppearanceImageMedia\")\n globalCharacterAppearancePreviousImgs GlobalCharacterAppearance[] @relation(\"GlobalCharacterAppearancePreviousImageMedia\")\n globalLocationImageImages GlobalLocationImage[] @relation(\"GlobalLocationImageMedia\")\n globalLocationImagePreviousImages GlobalLocationImage[] @relation(\"GlobalLocationImagePreviousImageMedia\")\n globalVoiceCustomVoices GlobalVoice[] @relation(\"GlobalVoiceCustomVoiceMedia\")\n\n @@index([createdAt])\n @@map(\"media_objects\")\n}\n\nmodel LegacyMediaRefBackup {\n id String @id @default(uuid())\n runId String\n tableName String\n rowId String\n fieldName String\n legacyValue String\n checksum String\n createdAt DateTime @default(now())\n\n @@index([runId])\n @@index([tableName, fieldName])\n @@map(\"legacy_media_refs_backup\")\n}\n"
},
{
"path": "scripts/billing-cleanup-pending-freezes.ts",
"content": "import { prisma } from '@/lib/prisma'\nimport { toMoneyNumber } from '@/lib/billing/money'\n\ntype CleanupStats = {\n scanned: number\n stale: number\n rolledBack: number\n skipped: number\n errors: number\n}\n\nfunction hasApplyFlag() {\n return process.argv.includes('--apply')\n}\n\nfunction parseHoursArg(defaultHours: number) {\n const arg = process.argv.find((item) => item.startsWith('--hours='))\n if (!arg) return defaultHours\n const value = Number(arg.slice('--hours='.length))\n if (!Number.isFinite(value) || value <= 0) return defaultHours\n return Math.floor(value)\n}\n\nfunction writeJson(payload: unknown) {\n process.stdout.write(`${JSON.stringify(payload, null, 2)}\\n`)\n}\n\nfunction writeError(payload: unknown) {\n process.stderr.write(`${typeof payload === 'string' ? payload : JSON.stringify(payload, null, 2)}\\n`)\n}\n\nasync function main() {\n const apply = hasApplyFlag()\n const hours = parseHoursArg(24)\n const cutoff = new Date(Date.now() - hours * 60 * 60 * 1000)\n\n const pending = await prisma.balanceFreeze.findMany({\n where: {\n status: 'pending',\n createdAt: { lt: cutoff },\n },\n orderBy: { createdAt: 'asc' },\n })\n\n const stats: CleanupStats = {\n scanned: pending.length,\n stale: pending.length,\n rolledBack: 0,\n skipped: 0,\n errors: 0,\n }\n\n if (!apply) {\n writeJson({\n mode: 'dry-run',\n hours,\n cutoff: cutoff.toISOString(),\n stalePendingCount: pending.length,\n stalePending: pending.map((f) => ({\n id: f.id,\n userId: f.userId,\n amount: toMoneyNumber(f.amount),\n createdAt: f.createdAt.toISOString(),\n })),\n })\n return\n }\n\n for (const freeze of pending) {\n try {\n await prisma.$transaction(async (tx) => {\n const current = await tx.balanceFreeze.findUnique({\n where: { id: freeze.id },\n })\n if (!current || current.status !== 'pending') {\n stats.skipped += 1\n return\n }\n\n const balance = await tx.userBalance.findUnique({\n where: { userId: current.userId },\n })\n if (!balance) {\n stats.skipped += 1\n return\n }\n\n const frozenAmount = toMoneyNumber(balance.frozenAmount)\n const freezeAmount = toMoneyNumber(current.amount)\n const nextFrozenAmount = Math.max(0, frozenAmount - freezeAmount)\n const frozenDelta = frozenAmount - nextFrozenAmount\n const balanceIncrement = frozenDelta\n\n await tx.userBalance.update({\n where: { userId: current.userId },\n data: {\n balance: { increment: balanceIncrement },\n frozenAmount: { decrement: frozenDelta },\n },\n })\n\n await tx.balanceFreeze.update({\n where: { id: current.id },\n data: {\n status: 'rolled_back',\n },\n })\n })\n stats.rolledBack += 1\n } catch (error) {\n stats.errors += 1\n writeError({\n tag: 'billing-cleanup-pending-freezes.rollback_failed',\n freezeId: freeze.id,\n userId: freeze.userId,\n amount: toMoneyNumber(freeze.amount),\n error: error instanceof Error ? error.message : String(error),\n })\n }\n }\n\n writeJson({\n mode: 'apply',\n hours,\n cutoff: cutoff.toISOString(),\n stats,\n })\n}\n\nmain()\n .catch((error) => {\n writeError({\n tag: 'billing-cleanup-pending-freezes.fatal',\n error: error instanceof Error ? error.message : String(error),\n })\n process.exit(1)\n })\n .finally(async () => {\n await prisma.$disconnect()\n })\n"
},
{
"path": "scripts/billing-reconcile-ledger.ts",
"content": "import { prisma } from '@/lib/prisma'\nimport { roundMoney, toMoneyNumber } from '@/lib/billing/money'\n\ntype UserLedgerRow = {\n userId: string\n balance: number\n frozenAmount: number\n txNetAmount: number\n ledgerAmount: number\n diff: number\n}\n\nfunction hasStrictFlag() {\n return process.argv.includes('--strict')\n}\n\nfunction write(payload: unknown) {\n process.stdout.write(`${JSON.stringify(payload, null, 2)}\\n`)\n}\n\nasync function main() {\n const strict = hasStrictFlag()\n\n const [balances, txByUser, pendingFreezes] = await Promise.all([\n prisma.userBalance.findMany({\n select: {\n userId: true,\n balance: true,\n frozenAmount: true,\n },\n }),\n prisma.balanceTransaction.groupBy({\n by: ['userId'],\n _sum: { amount: true },\n }),\n prisma.balanceFreeze.findMany({\n where: { status: 'pending' },\n select: {\n id: true,\n userId: true,\n taskId: true,\n amount: true,\n createdAt: true,\n },\n orderBy: { createdAt: 'asc' },\n }),\n ])\n\n const txNetByUser = new Map()\n for (const row of txByUser) {\n txNetByUser.set(row.userId, roundMoney(toMoneyNumber(row._sum.amount), 8))\n }\n\n const ledgerRows: UserLedgerRow[] = balances.map((row) => {\n const balance = toMoneyNumber(row.balance)\n const frozenAmount = toMoneyNumber(row.frozenAmount)\n const txNetAmount = roundMoney(txNetByUser.get(row.userId) || 0, 8)\n const ledgerAmount = roundMoney(balance + frozenAmount, 8)\n return {\n userId: row.userId,\n balance,\n frozenAmount,\n txNetAmount,\n ledgerAmount,\n diff: roundMoney(ledgerAmount - txNetAmount, 8),\n }\n })\n\n const nonZeroDiffUsers = ledgerRows.filter((row) => Math.abs(row.diff) > 1e-8)\n\n const pendingTaskIds = pendingFreezes\n .map((row) => row.taskId)\n .filter((taskId): taskId is string => typeof taskId === 'string' && taskId.length > 0)\n const tasks = pendingTaskIds.length > 0\n ? await prisma.task.findMany({\n where: { id: { in: pendingTaskIds } },\n select: { id: true, status: true },\n })\n : []\n const taskStatusById = new Map(tasks.map((row) => [row.id, row.status]))\n const activeStatuses = new Set(['queued', 'processing'])\n const orphanPendingFreezes = pendingFreezes.filter((freeze) => {\n if (!freeze.taskId) return true\n const status = taskStatusById.get(freeze.taskId)\n if (!status) return true\n return !activeStatuses.has(status)\n })\n\n const result = {\n strict,\n checkedAt: new Date().toISOString(),\n totals: {\n users: balances.length,\n txUsers: txByUser.length,\n pendingFreezes: pendingFreezes.length,\n nonZeroDiffUsers: nonZeroDiffUsers.length,\n orphanPendingFreezes: orphanPendingFreezes.length,\n },\n nonZeroDiffUsers,\n orphanPendingFreezes: orphanPendingFreezes.map((row) => ({\n id: row.id,\n userId: row.userId,\n taskId: row.taskId,\n amount: toMoneyNumber(row.amount),\n createdAt: row.createdAt.toISOString(),\n })),\n }\n\n write(result)\n\n if (strict && (nonZeroDiffUsers.length > 0 || orphanPendingFreezes.length > 0)) {\n process.exitCode = 1\n }\n}\n\nmain()\n .catch((error) => {\n write({\n error: error instanceof Error ? error.message : String(error),\n })\n process.exitCode = 1\n })\n .finally(async () => {\n await prisma.$disconnect()\n })\n"
},
{
"path": "scripts/bull-board.ts",
"content": "import { createScopedLogger } from '@/lib/logging/core'\nimport express, { type NextFunction, type Request, type Response } from 'express'\nimport { createBullBoard } from '@bull-board/api'\nimport { BullMQAdapter } from '@bull-board/api/bullMQAdapter'\nimport { ExpressAdapter } from '@bull-board/express'\nimport { imageQueue, textQueue, videoQueue, voiceQueue } from '@/lib/task/queues'\n\nconst host = process.env.BULL_BOARD_HOST || '127.0.0.1'\nconst port = Number.parseInt(process.env.BULL_BOARD_PORT || '3010', 10) || 3010\nconst basePath = process.env.BULL_BOARD_BASE_PATH || '/admin/queues'\nconst authUser = process.env.BULL_BOARD_USER\nconst authPassword = process.env.BULL_BOARD_PASSWORD\nconst logger = createScopedLogger({\n module: 'ops.bull_board',\n})\n\nfunction unauthorized(res: Response) {\n res.setHeader('WWW-Authenticate', 'Basic realm=\"BullMQ Board\"')\n res.status(401).send('Authentication required')\n}\n\nfunction basicAuthMiddleware(req: Request, res: Response, next: NextFunction) {\n if (!authUser && !authPassword) {\n next()\n return\n }\n\n const authorization = req.headers.authorization\n if (!authorization?.startsWith('Basic ')) {\n unauthorized(res)\n return\n }\n\n const encoded = authorization.slice(6).trim()\n let decoded = ''\n\n try {\n decoded = Buffer.from(encoded, 'base64').toString('utf8')\n } catch {\n unauthorized(res)\n return\n }\n\n const index = decoded.indexOf(':')\n if (index === -1) {\n unauthorized(res)\n return\n }\n\n const username = decoded.slice(0, index)\n const password = decoded.slice(index + 1)\n if (username !== (authUser || '') || password !== (authPassword || '')) {\n unauthorized(res)\n return\n }\n\n next()\n}\n\nconst serverAdapter = new ExpressAdapter()\nserverAdapter.setBasePath(basePath)\n\ncreateBullBoard({\n queues: [\n new BullMQAdapter(imageQueue),\n new BullMQAdapter(videoQueue),\n new BullMQAdapter(voiceQueue),\n new BullMQAdapter(textQueue),\n ],\n serverAdapter,\n})\n\nconst app = express()\napp.disable('x-powered-by')\napp.use(basePath, basicAuthMiddleware, serverAdapter.getRouter())\n\nconst server = app.listen(port, host, () => {\n const secured = authUser || authPassword ? 'enabled' : 'disabled'\n logger.info({\n action: 'bull_board.started',\n message: 'bull board listening',\n details: {\n host,\n port,\n basePath,\n auth: secured,\n },\n })\n})\n\nasync function shutdown(signal: string) {\n logger.info({\n action: 'bull_board.shutdown',\n message: 'bull board shutting down',\n details: {\n signal,\n },\n })\n await Promise.allSettled([imageQueue.close(), videoQueue.close(), voiceQueue.close(), textQueue.close()])\n await new Promise((resolve) => server.close(() => resolve()))\n process.exit(0)\n}\n\nprocess.on('SIGINT', () => void shutdown('SIGINT'))\nprocess.on('SIGTERM', () => void shutdown('SIGTERM'))\n"
},
{
"path": "scripts/check-api-handler.ts",
"content": "import { logInfo as _ulogInfo, logError as _ulogError } from '@/lib/logging/core'\nimport { execSync } from 'node:child_process'\n\nconst ALLOWLIST = new Set([\n 'src/app/api/auth/[...nextauth]/route.ts',\n 'src/app/api/files/[...path]/route.ts',\n 'src/app/api/system/boot-id/route.ts',\n])\n\nfunction main() {\n const output = execSync(\"rg --files src/app/api | rg 'route\\\\.ts$'\", { encoding: 'utf8' })\n const files = output\n .split('\\n')\n .map((line) => line.trim())\n .filter(Boolean)\n\n const missing: string[] = []\n\n for (const file of files) {\n if (ALLOWLIST.has(file)) continue\n const hasApiHandler = execSync(`rg -n \\\"apiHandler\\\" ${JSON.stringify(file)} || true`, { encoding: 'utf8' }).trim().length > 0\n if (!hasApiHandler) {\n missing.push(file)\n }\n }\n\n if (missing.length > 0) {\n _ulogError('[check-api-handler] missing apiHandler in:')\n for (const file of missing) {\n _ulogError(`- ${file}`)\n }\n process.exit(1)\n }\n\n _ulogInfo(`[check-api-handler] ok total=${files.length} allowlist=${ALLOWLIST.size}`)\n}\n\nmain()\n"
},
{
"path": "scripts/check-capability-catalog.mjs",
"content": "import { promises as fs } from 'node:fs'\nimport path from 'node:path'\n\n const CATALOG_DIR = path.resolve(process.cwd(), 'standards/capabilities')\nconst CAPABILITY_NAMESPACES = new Set(['llm', 'image', 'video', 'audio', 'lipsync'])\nconst CAPABILITY_NAMESPACE_ALLOWED_FIELDS = {\n llm: new Set(['reasoningEffortOptions', 'fieldI18n']),\n image: new Set(['resolutionOptions', 'fieldI18n']),\n video: new Set([\n 'generationModeOptions',\n 'generateAudioOptions',\n 'durationOptions',\n 'fpsOptions',\n 'resolutionOptions',\n 'firstlastframe',\n 'supportGenerateAudio',\n 'fieldI18n',\n ]),\n audio: new Set(['voiceOptions', 'rateOptions', 'fieldI18n']),\n lipsync: new Set(['modeOptions', 'fieldI18n']),\n}\nconst CAPABILITY_NAMESPACE_I18N_FIELDS = {\n llm: { reasoningEffort: 'reasoningEffortOptions' },\n image: { resolution: 'resolutionOptions' },\n video: {\n generationMode: 'generationModeOptions',\n generateAudio: 'generateAudioOptions',\n duration: 'durationOptions',\n fps: 'fpsOptions',\n resolution: 'resolutionOptions',\n },\n audio: { voice: 'voiceOptions', rate: 'rateOptions' },\n lipsync: { mode: 'modeOptions' },\n}\nconst MODEL_TYPES = new Set(['llm', 'image', 'video', 'audio', 'lipsync'])\n\nfunction isRecord(value) {\n return !!value && typeof value === 'object' && !Array.isArray(value)\n}\n\nfunction isNonEmptyString(value) {\n return typeof value === 'string' && value.trim().length > 0\n}\n\nfunction isI18nKey(value) {\n return isNonEmptyString(value) && value.includes('.')\n}\n\nfunction isStringArray(value) {\n return Array.isArray(value) && value.every((item) => isNonEmptyString(item))\n}\n\nfunction isNumberArray(value) {\n return Array.isArray(value) && value.every((item) => typeof item === 'number' && Number.isFinite(item))\n}\n\nfunction isBooleanArray(value) {\n return Array.isArray(value) && value.every((item) => typeof item === 'boolean')\n}\n\nfunction parseModelKeyStrict(value) {\n if (!isNonEmptyString(value)) return null\n const raw = value.trim()\n const marker = raw.indexOf('::')\n if (marker === -1) return null\n const provider = raw.slice(0, marker).trim()\n const modelId = raw.slice(marker + 2).trim()\n if (!provider || !modelId) return null\n return { provider, modelId, modelKey: `${provider}::${modelId}` }\n}\n\nfunction pushIssue(issues, file, index, field, message) {\n issues.push({ file, index, field, message })\n}\n\nfunction validateAllowedFields(issues, file, index, namespace, namespaceValue) {\n if (!isRecord(namespaceValue)) return\n const allowedFields = CAPABILITY_NAMESPACE_ALLOWED_FIELDS[namespace]\n for (const field of Object.keys(namespaceValue)) {\n if (allowedFields.has(field)) continue\n if (field === 'i18n') {\n pushIssue(issues, file, index, `capabilities.${namespace}.${field}`, 'use fieldI18n instead of i18n')\n continue\n }\n pushIssue(issues, file, index, `capabilities.${namespace}.${field}`, `unknown capability field: ${field}`)\n }\n}\n\nfunction validateFieldI18nMap(issues, file, index, namespace, namespaceValue) {\n if (!isRecord(namespaceValue)) return\n if (namespaceValue.fieldI18n === undefined) return\n if (!isRecord(namespaceValue.fieldI18n)) {\n pushIssue(issues, file, index, `capabilities.${namespace}.fieldI18n`, 'fieldI18n must be an object')\n return\n }\n\n const allowedI18nFields = CAPABILITY_NAMESPACE_I18N_FIELDS[namespace]\n for (const [fieldName, fieldConfig] of Object.entries(namespaceValue.fieldI18n)) {\n if (!(fieldName in allowedI18nFields)) {\n pushIssue(issues, file, index, `capabilities.${namespace}.fieldI18n.${fieldName}`, `unknown i18n field: ${fieldName}`)\n continue\n }\n if (!isRecord(fieldConfig)) {\n pushIssue(issues, file, index, `capabilities.${namespace}.fieldI18n.${fieldName}`, 'field i18n config must be an object')\n continue\n }\n\n if (fieldConfig.labelKey !== undefined && !isI18nKey(fieldConfig.labelKey)) {\n pushIssue(issues, file, index, `capabilities.${namespace}.fieldI18n.${fieldName}.labelKey`, 'labelKey must be an i18n key')\n }\n if (fieldConfig.unitKey !== undefined && !isI18nKey(fieldConfig.unitKey)) {\n pushIssue(issues, file, index, `capabilities.${namespace}.fieldI18n.${fieldName}.unitKey`, 'unitKey must be an i18n key')\n }\n if (fieldConfig.optionLabelKeys !== undefined) {\n if (!isRecord(fieldConfig.optionLabelKeys)) {\n pushIssue(\n issues,\n file,\n index,\n `capabilities.${namespace}.fieldI18n.${fieldName}.optionLabelKeys`,\n 'optionLabelKeys must be an object',\n )\n continue\n }\n const optionFieldName = allowedI18nFields[fieldName]\n const optionsRaw = namespaceValue[optionFieldName]\n const allowedOptions = Array.isArray(optionsRaw) ? new Set(optionsRaw.map((value) => String(value))) : null\n for (const [optionValue, optionLabel] of Object.entries(fieldConfig.optionLabelKeys)) {\n if (!isI18nKey(optionLabel)) {\n pushIssue(\n issues,\n file,\n index,\n `capabilities.${namespace}.fieldI18n.${fieldName}.optionLabelKeys.${optionValue}`,\n 'option label must be an i18n key',\n )\n }\n if (allowedOptions && !allowedOptions.has(optionValue)) {\n pushIssue(\n issues,\n file,\n index,\n `capabilities.${namespace}.fieldI18n.${fieldName}.optionLabelKeys.${optionValue}`,\n `option ${optionValue} is not defined in ${optionFieldName}`,\n )\n }\n }\n }\n }\n}\n\nfunction validateCapabilitiesForModelType(issues, file, index, modelType, capabilities) {\n if (capabilities === undefined || capabilities === null) return\n if (!isRecord(capabilities)) {\n pushIssue(issues, file, index, 'capabilities', 'capabilities must be an object')\n return\n }\n\n const expectedNamespace = modelType\n for (const namespace of Object.keys(capabilities)) {\n if (!CAPABILITY_NAMESPACES.has(namespace)) {\n pushIssue(issues, file, index, `capabilities.${namespace}`, `unknown capabilities namespace: ${namespace}`)\n continue\n }\n if (namespace !== expectedNamespace) {\n pushIssue(\n issues,\n file,\n index,\n `capabilities.${namespace}`,\n `namespace ${namespace} is not allowed for model type ${modelType}`,\n )\n }\n }\n\n const llm = capabilities.llm\n if (llm !== undefined) {\n if (!isRecord(llm)) {\n pushIssue(issues, file, index, 'capabilities.llm', 'llm capabilities must be an object')\n } else {\n validateAllowedFields(issues, file, index, 'llm', llm)\n if (llm.reasoningEffortOptions !== undefined && !isStringArray(llm.reasoningEffortOptions)) {\n pushIssue(issues, file, index, 'capabilities.llm.reasoningEffortOptions', 'must be string array')\n }\n validateFieldI18nMap(issues, file, index, 'llm', llm)\n }\n }\n\n const image = capabilities.image\n if (image !== undefined) {\n if (!isRecord(image)) {\n pushIssue(issues, file, index, 'capabilities.image', 'image capabilities must be an object')\n } else {\n validateAllowedFields(issues, file, index, 'image', image)\n if (image.resolutionOptions !== undefined && !isStringArray(image.resolutionOptions)) {\n pushIssue(issues, file, index, 'capabilities.image.resolutionOptions', 'must be string array')\n }\n validateFieldI18nMap(issues, file, index, 'image', image)\n }\n }\n\n const video = capabilities.video\n if (video !== undefined) {\n if (!isRecord(video)) {\n pushIssue(issues, file, index, 'capabilities.video', 'video capabilities must be an object')\n } else {\n validateAllowedFields(issues, file, index, 'video', video)\n if (video.generationModeOptions !== undefined && !isStringArray(video.generationModeOptions)) {\n pushIssue(issues, file, index, 'capabilities.video.generationModeOptions', 'must be string array')\n }\n if (video.generateAudioOptions !== undefined && !isBooleanArray(video.generateAudioOptions)) {\n pushIssue(issues, file, index, 'capabilities.video.generateAudioOptions', 'must be boolean array')\n }\n if (video.durationOptions !== undefined && !isNumberArray(video.durationOptions)) {\n pushIssue(issues, file, index, 'capabilities.video.durationOptions', 'must be number array')\n }\n if (video.fpsOptions !== undefined && !isNumberArray(video.fpsOptions)) {\n pushIssue(issues, file, index, 'capabilities.video.fpsOptions', 'must be number array')\n }\n if (video.resolutionOptions !== undefined && !isStringArray(video.resolutionOptions)) {\n pushIssue(issues, file, index, 'capabilities.video.resolutionOptions', 'must be string array')\n }\n if (video.supportGenerateAudio !== undefined && typeof video.supportGenerateAudio !== 'boolean') {\n pushIssue(issues, file, index, 'capabilities.video.supportGenerateAudio', 'must be boolean')\n }\n if (video.firstlastframe !== undefined && typeof video.firstlastframe !== 'boolean') {\n pushIssue(issues, file, index, 'capabilities.video.firstlastframe', 'must be boolean')\n }\n validateFieldI18nMap(issues, file, index, 'video', video)\n }\n }\n\n const audio = capabilities.audio\n if (audio !== undefined) {\n if (!isRecord(audio)) {\n pushIssue(issues, file, index, 'capabilities.audio', 'audio capabilities must be an object')\n } else {\n validateAllowedFields(issues, file, index, 'audio', audio)\n if (audio.voiceOptions !== undefined && !isStringArray(audio.voiceOptions)) {\n pushIssue(issues, file, index, 'capabilities.audio.voiceOptions', 'must be string array')\n }\n if (audio.rateOptions !== undefined && !isStringArray(audio.rateOptions)) {\n pushIssue(issues, file, index, 'capabilities.audio.rateOptions', 'must be string array')\n }\n validateFieldI18nMap(issues, file, index, 'audio', audio)\n }\n }\n\n const lipsync = capabilities.lipsync\n if (lipsync !== undefined) {\n if (!isRecord(lipsync)) {\n pushIssue(issues, file, index, 'capabilities.lipsync', 'lipsync capabilities must be an object')\n } else {\n validateAllowedFields(issues, file, index, 'lipsync', lipsync)\n if (lipsync.modeOptions !== undefined && !isStringArray(lipsync.modeOptions)) {\n pushIssue(issues, file, index, 'capabilities.lipsync.modeOptions', 'must be string array')\n }\n validateFieldI18nMap(issues, file, index, 'lipsync', lipsync)\n }\n }\n}\n\nasync function listCatalogFiles() {\n const entries = await fs.readdir(CATALOG_DIR, { withFileTypes: true })\n return entries\n .filter((entry) => entry.isFile() && entry.name.endsWith('.json'))\n .map((entry) => path.join(CATALOG_DIR, entry.name))\n}\n\nasync function readCatalog(filePath) {\n const raw = await fs.readFile(filePath, 'utf8')\n const parsed = JSON.parse(raw)\n if (!Array.isArray(parsed)) {\n throw new Error(`catalog must be an array: ${filePath}`)\n }\n return parsed\n}\n\nasync function main() {\n const issues = []\n const files = await listCatalogFiles()\n if (files.length === 0) {\n throw new Error(`no catalog files found in ${CATALOG_DIR}`)\n }\n\n for (const filePath of files) {\n const catalogItems = await readCatalog(filePath)\n for (let index = 0; index < catalogItems.length; index += 1) {\n const item = catalogItems[index]\n if (!isRecord(item)) {\n pushIssue(issues, filePath, index, 'entry', 'entry must be an object')\n continue\n }\n\n if (!isNonEmptyString(item.modelType) || !MODEL_TYPES.has(item.modelType)) {\n pushIssue(issues, filePath, index, 'modelType', 'modelType must be llm/image/video/audio/lipsync')\n continue\n }\n\n if (!isNonEmptyString(item.provider)) {\n pushIssue(issues, filePath, index, 'provider', 'provider must be a non-empty string')\n }\n if (!isNonEmptyString(item.modelId)) {\n pushIssue(issues, filePath, index, 'modelId', 'modelId must be a non-empty string')\n }\n\n const modelKey = `${item.provider || ''}::${item.modelId || ''}`\n if (!parseModelKeyStrict(modelKey)) {\n pushIssue(issues, filePath, index, 'modelKey', 'provider/modelId must compose a valid provider::modelId')\n }\n\n validateCapabilitiesForModelType(issues, filePath, index, item.modelType, item.capabilities)\n }\n }\n\n if (issues.length === 0) {\n process.stdout.write(`[check-capability-catalog] OK (${files.length} files)\\n`)\n return\n }\n\n const maxPrint = 50\n for (const issue of issues.slice(0, maxPrint)) {\n process.stdout.write(`[check-capability-catalog] ${issue.file}#${issue.index} ${issue.field}: ${issue.message}\\n`)\n }\n if (issues.length > maxPrint) {\n process.stdout.write(`[check-capability-catalog] ... ${issues.length - maxPrint} more issues\\n`)\n }\n process.exitCode = 1\n}\n\nmain().catch((error) => {\n process.stderr.write(`[check-capability-catalog] failed: ${String(error)}\\n`)\n process.exitCode = 1\n})\n"
},
{
"path": "scripts/check-image-urls-contract.ts",
"content": "import { logInfo as _ulogInfo, logError as _ulogError } from '@/lib/logging/core'\nimport { prisma } from '@/lib/prisma'\nimport { decodeImageUrlsFromDb } from '@/lib/contracts/image-urls-contract'\n\ntype AppearanceRow = {\n id: string\n imageUrls: string | null\n previousImageUrls: string | null\n}\n\ntype DynamicModel = {\n findMany: (args: unknown) => Promise\n}\n\nconst BATCH_SIZE = 500\n\nconst MODELS: Array<{ name: string; model: string }> = [\n { name: 'CharacterAppearance', model: 'characterAppearance' },\n { name: 'GlobalCharacterAppearance', model: 'globalCharacterAppearance' },\n]\n\nconst prismaDynamic = prisma as unknown as Record\n\nfunction print(message: string) {\n process.stdout.write(`${message}\\n`)\n}\n\nasync function checkModel(modelName: string, modelKey: string) {\n const model = prismaDynamic[modelKey]\n if (!model) {\n throw new Error(`Prisma model not found: ${modelKey}`)\n }\n\n let scanned = 0\n let violations = 0\n const samples: Array<{ id: string; field: 'imageUrls' | 'previousImageUrls'; message: string; value: string | null }> = []\n let cursor: string | null = null\n\n while (true) {\n const rows = await model.findMany({\n select: {\n id: true,\n imageUrls: true,\n previousImageUrls: true,\n },\n ...(cursor\n ? {\n cursor: { id: cursor },\n skip: 1,\n }\n : {}),\n orderBy: { id: 'asc' },\n take: BATCH_SIZE,\n })\n\n if (rows.length === 0) break\n\n for (const row of rows) {\n scanned += 1\n\n for (const fieldName of ['imageUrls', 'previousImageUrls'] as const) {\n try {\n decodeImageUrlsFromDb(row[fieldName], `${modelName}.${fieldName}`)\n } catch (error) {\n violations += 1\n if (samples.length < 20) {\n samples.push({\n id: row.id,\n field: fieldName,\n message: error instanceof Error ? error.message : String(error),\n value: row[fieldName],\n })\n }\n }\n }\n }\n\n cursor = rows[rows.length - 1]?.id || null\n }\n\n const summary = `[check-image-urls-contract] ${modelName}: scanned=${scanned} violations=${violations}`\n _ulogInfo(summary)\n print(summary)\n if (samples.length > 0) {\n _ulogError(`[check-image-urls-contract] ${modelName}: samples=${JSON.stringify(samples, null, 2)}`)\n }\n\n return { scanned, violations }\n}\n\nasync function main() {\n let totalScanned = 0\n let totalViolations = 0\n\n for (const target of MODELS) {\n const result = await checkModel(target.name, target.model)\n totalScanned += result.scanned\n totalViolations += result.violations\n }\n\n if (totalViolations > 0) {\n _ulogError(`[check-image-urls-contract] failed scanned=${totalScanned} violations=${totalViolations}`)\n print(`[check-image-urls-contract] failed scanned=${totalScanned} violations=${totalViolations}`)\n process.exitCode = 1\n return\n }\n\n print(`[check-image-urls-contract] ok scanned=${totalScanned}`)\n}\n\nmain()\n .catch((error) => {\n _ulogError('[check-image-urls-contract] failed:', error)\n process.exitCode = 1\n })\n .finally(async () => {\n await prisma.$disconnect()\n })\n"
},
{
"path": "scripts/check-log-semantic.ts",
"content": "import fs from 'node:fs'\n\ntype Rule = {\n file: string\n patterns: string[]\n}\n\nconst RULES: Rule[] = [\n {\n file: 'src/lib/api-errors.ts',\n patterns: ['x-request-id', 'api.request.start', 'api.request.finish', 'api.request.error'],\n },\n {\n file: 'src/lib/workers/shared.ts',\n patterns: ['worker.start', 'worker.completed', 'worker.failed', 'durationMs', 'errorCode'],\n },\n {\n file: 'src/app/api/sse/route.ts',\n patterns: ['sse.connect', 'sse.replay', 'sse.disconnect'],\n },\n {\n file: 'scripts/watchdog.ts',\n patterns: ['watchdog.started', 'watchdog.tick.ok', 'watchdog.tick.failed'],\n },\n {\n file: 'scripts/bull-board.ts',\n patterns: ['bull_board.started', 'bull_board.shutdown'],\n },\n {\n file: 'src/lib/task/submitter.ts',\n patterns: ['requestId', 'task.submit.created', 'task.submit.enqueued'],\n },\n {\n file: 'src/lib/task/types.ts',\n patterns: ['trace', 'requestId'],\n },\n]\n\nfunction read(file: string) {\n return fs.readFileSync(file, 'utf8')\n}\n\nfunction checkRules() {\n const violations: string[] = []\n for (const rule of RULES) {\n const content = read(rule.file)\n for (const pattern of rule.patterns) {\n if (!content.includes(pattern)) {\n violations.push(`${rule.file} missing \"${pattern}\"`)\n }\n }\n }\n return violations\n}\n\nfunction checkSubmitTaskRoutes() {\n const root = 'src/app/api'\n const files = walk(root).filter((file) => file.endsWith('/route.ts'))\n const submitTaskFiles = files.filter((file) => read(file).includes('submitTask('))\n const violations: string[] = []\n\n for (const file of submitTaskFiles) {\n const content = read(file)\n if (!content.includes('getRequestId')) {\n violations.push(`${file} uses submitTask but does not import getRequestId`)\n continue\n }\n if (!content.includes('requestId: getRequestId(request)')) {\n violations.push(`${file} uses submitTask but does not pass requestId`)\n }\n }\n\n return { submitTaskFiles, violations }\n}\n\nfunction walk(dir: string): string[] {\n const entries = fs.readdirSync(dir, { withFileTypes: true })\n const out: string[] = []\n\n for (const entry of entries) {\n const next = `${dir}/${entry.name}`\n if (entry.isDirectory()) {\n out.push(...walk(next))\n } else {\n out.push(next)\n }\n }\n\n return out\n}\n\nfunction main() {\n const violations = checkRules()\n const submitTaskResult = checkSubmitTaskRoutes()\n violations.push(...submitTaskResult.violations)\n\n if (violations.length > 0) {\n process.stderr.write('[check:log-semantic] semantic violations detected:\\n')\n for (const violation of violations) {\n process.stderr.write(`- ${violation}\\n`)\n }\n process.exit(1)\n }\n\n process.stdout.write(\n `[check:log-semantic] ok rules=${RULES.length} submitTaskRoutes=${submitTaskResult.submitTaskFiles.length}\\n`,\n )\n}\n\nmain()\n"
},
{
"path": "scripts/check-media-normalization.ts",
"content": "import { execSync } from 'node:child_process'\n\nconst TARGETS = ['src/app/api', 'src/lib']\n\nconst EXTRACT_ALLOWLIST = new Set([\n 'src/lib/media/service.ts',\n 'src/lib/voice/generate-voice-line.ts',\n])\n\nconst FETCH_MEDIA_ALLOWLIST = new Set([\n 'src/lib/media-process.ts',\n 'src/lib/image-cache.ts',\n 'src/lib/image-label.ts',\n 'src/lib/workers/utils.ts',\n 'src/app/api/novel-promotion/[projectId]/download-images/route.ts',\n 'src/app/api/novel-promotion/[projectId]/download-videos/route.ts',\n 'src/app/api/novel-promotion/[projectId]/download-voices/route.ts',\n 'src/app/api/novel-promotion/[projectId]/update-asset-label/route.ts',\n 'src/app/api/novel-promotion/[projectId]/voice-generate/route.ts',\n 'src/app/api/novel-promotion/[projectId]/video-proxy/route.ts',\n])\n\nfunction run(cmd: string): string {\n try {\n return execSync(cmd, { encoding: 'utf8' })\n } catch (error: unknown) {\n if (error && typeof error === 'object' && 'stdout' in error) {\n const stdout = (error as { stdout?: unknown }).stdout\n return typeof stdout === 'string' ? stdout : ''\n }\n return ''\n }\n}\n\nfunction parseLines(output: string): string[] {\n return output\n .split('\\n')\n .map((line) => line.trim())\n .filter(Boolean)\n}\n\nfunction getFile(line: string): string {\n return line.split(':', 1)[0] || ''\n}\n\nfunction getCode(line: string): string {\n const parts = line.split(':')\n return parts.slice(2).join(':').trim()\n}\n\nfunction extractFetchArg(code: string): string {\n const matched = code.match(/fetch\\(\\s*([^)]+)\\)/)\n return matched?.[1]?.trim() || ''\n}\n\nfunction isSafeFetchArg(arg: string): boolean {\n if (!arg) return false\n if (/^toFetchableUrl\\(/.test(arg)) return true\n if (/^['\"`]/.test(arg)) return true\n if (/^new URL\\(/.test(arg)) return true\n return false\n}\n\nfunction isMediaLikeFetchArg(arg: string): boolean {\n return /(image|video|audio|signed).*url/i.test(arg) || /url.*(image|video|audio|signed)/i.test(arg)\n}\n\nfunction main() {\n const targetExpr = TARGETS.join(' ')\n\n // 规则 1：业务代码中不允许直接调用 extractStorageKey（统一走 resolveStorageKeyFromMediaValue）\n const extractOutput = run(`rg -n \"extractStorageKey\\\\\\\\(\" ${targetExpr}`)\n const extractLines = parseLines(extractOutput)\n const extractViolations = extractLines.filter((line) => {\n const file = getFile(line)\n if (file.startsWith('src/lib/storage/')) return false\n return !EXTRACT_ALLOWLIST.has(file)\n })\n\n // 规则 2：媒体相关 fetch 必须包裹 toFetchableUrl\n const fetchOutput = run(`rg -n \"fetch\\\\\\\\(\" ${targetExpr}`)\n const fetchLines = parseLines(fetchOutput)\n const fetchViolations = fetchLines.filter((line) => {\n const file = getFile(line)\n if (!FETCH_MEDIA_ALLOWLIST.has(file)) return false\n const code = getCode(line)\n const arg = extractFetchArg(code)\n if (!isMediaLikeFetchArg(arg)) return false\n return !isSafeFetchArg(arg)\n })\n\n const violations = [\n ...extractViolations.map((line) => `extractStorageKey forbidden: ${line}`),\n ...fetchViolations.map((line) => `fetch without toFetchableUrl: ${line}`),\n ]\n\n if (violations.length > 0) {\n process.stderr.write('[check:media-normalization] found violations:\\n')\n for (const item of violations) {\n process.stderr.write(`- ${item}\\n`)\n }\n process.exit(1)\n }\n\n process.stdout.write(\n `[check:media-normalization] ok extract_scanned=${extractLines.length} fetch_scanned=${fetchLines.length} allow_extract=${EXTRACT_ALLOWLIST.size} allow_fetch=${FETCH_MEDIA_ALLOWLIST.size}\\n`,\n )\n}\n\nmain()\n"
},
{
"path": "scripts/check-model-config-contract.mjs",
"content": "let prisma\n\nconst STRICT = process.argv.includes('--strict')\nconst MODEL_FIELDS = [\n 'analysisModel',\n 'characterModel',\n 'locationModel',\n 'storyboardModel',\n 'editModel',\n 'videoModel',\n]\nconst MAX_SAMPLES = 200\nconst CAPABILITY_NAMESPACES = new Set(['llm', 'image', 'video', 'audio', 'lipsync'])\nconst MODEL_TYPES = new Set(['llm', 'image', 'video', 'audio', 'lipsync'])\nconst CAPABILITY_NAMESPACE_ALLOWED_FIELDS = {\n llm: new Set(['reasoningEffortOptions', 'fieldI18n']),\n image: new Set(['resolutionOptions', 'fieldI18n']),\n video: new Set([\n 'durationOptions',\n 'fpsOptions',\n 'resolutionOptions',\n 'firstlastframe',\n 'supportGenerateAudio',\n 'fieldI18n',\n ]),\n audio: new Set(['voiceOptions', 'rateOptions', 'fieldI18n']),\n lipsync: new Set(['modeOptions', 'fieldI18n']),\n}\n\nconst CAPABILITY_NAMESPACE_I18N_FIELDS = {\n llm: {\n reasoningEffort: 'reasoningEffortOptions',\n },\n image: {\n resolution: 'resolutionOptions',\n },\n video: {\n duration: 'durationOptions',\n fps: 'fpsOptions',\n resolution: 'resolutionOptions',\n },\n audio: {\n voice: 'voiceOptions',\n rate: 'rateOptions',\n },\n lipsync: {\n mode: 'modeOptions',\n },\n}\n\nfunction isRecord(value) {\n return !!value && typeof value === 'object' && !Array.isArray(value)\n}\n\nfunction isNonEmptyString(value) {\n return typeof value === 'string' && value.trim().length > 0\n}\n\nfunction isStringArray(value) {\n return Array.isArray(value) && value.every((item) => isNonEmptyString(item))\n}\n\nfunction isNumberArray(value) {\n return Array.isArray(value) && value.every((item) => typeof item === 'number' && Number.isFinite(item))\n}\n\nfunction parseModelKeyStrict(value) {\n if (!isNonEmptyString(value)) return null\n const raw = value.trim()\n const marker = raw.indexOf('::')\n if (marker === -1) return null\n const provider = raw.slice(0, marker).trim()\n const modelId = raw.slice(marker + 2).trim()\n if (!provider || !modelId) return null\n return {\n provider,\n modelId,\n modelKey: `${provider}::${modelId}`,\n }\n}\n\nfunction addSample(summary, sample) {\n if (summary.samples.length >= MAX_SAMPLES) return\n summary.samples.push(sample)\n}\n\nfunction pushIssue(issues, field, message) {\n issues.push({ field, message })\n}\n\nfunction isI18nKey(value) {\n return isNonEmptyString(value) && value.includes('.')\n}\n\nfunction validateAllowedFields(issues, namespace, namespaceValue) {\n if (!isRecord(namespaceValue)) return\n const allowedFields = CAPABILITY_NAMESPACE_ALLOWED_FIELDS[namespace]\n for (const field of Object.keys(namespaceValue)) {\n if (allowedFields.has(field)) continue\n if (field === 'i18n') {\n pushIssue(issues, `capabilities.${namespace}.${field}`, 'use fieldI18n instead of i18n')\n continue\n }\n pushIssue(issues, `capabilities.${namespace}.${field}`, `unknown capability field: ${field}`)\n }\n}\n\nfunction validateFieldI18nMap(issues, namespace, namespaceValue) {\n if (!isRecord(namespaceValue)) return\n if (namespaceValue.fieldI18n === undefined) return\n if (!isRecord(namespaceValue.fieldI18n)) {\n pushIssue(issues, `capabilities.${namespace}.fieldI18n`, 'fieldI18n must be an object')\n return\n }\n\n const allowedI18nFields = CAPABILITY_NAMESPACE_I18N_FIELDS[namespace]\n for (const [fieldName, fieldConfig] of Object.entries(namespaceValue.fieldI18n)) {\n if (!(fieldName in allowedI18nFields)) {\n pushIssue(\n issues,\n `capabilities.${namespace}.fieldI18n.${fieldName}`,\n `unknown i18n field: ${fieldName}`,\n )\n continue\n }\n if (!isRecord(fieldConfig)) {\n pushIssue(\n issues,\n `capabilities.${namespace}.fieldI18n.${fieldName}`,\n 'field i18n config must be an object',\n )\n continue\n }\n\n if (fieldConfig.labelKey !== undefined && !isI18nKey(fieldConfig.labelKey)) {\n pushIssue(\n issues,\n `capabilities.${namespace}.fieldI18n.${fieldName}.labelKey`,\n 'labelKey must be an i18n key',\n )\n }\n if (fieldConfig.unitKey !== undefined && !isI18nKey(fieldConfig.unitKey)) {\n pushIssue(\n issues,\n `capabilities.${namespace}.fieldI18n.${fieldName}.unitKey`,\n 'unitKey must be an i18n key',\n )\n }\n if (fieldConfig.optionLabelKeys !== undefined) {\n if (!isRecord(fieldConfig.optionLabelKeys)) {\n pushIssue(\n issues,\n `capabilities.${namespace}.fieldI18n.${fieldName}.optionLabelKeys`,\n 'optionLabelKeys must be an object',\n )\n continue\n }\n\n const optionFieldName = allowedI18nFields[fieldName]\n const allowedOptionsRaw = namespaceValue[optionFieldName]\n const allowedOptions = Array.isArray(allowedOptionsRaw)\n ? new Set(allowedOptionsRaw.map((value) => String(value)))\n : null\n\n for (const [optionValue, optionLabelKey] of Object.entries(fieldConfig.optionLabelKeys)) {\n if (!isI18nKey(optionLabelKey)) {\n pushIssue(\n issues,\n `capabilities.${namespace}.fieldI18n.${fieldName}.optionLabelKeys.${optionValue}`,\n 'option label must be an i18n key',\n )\n }\n if (allowedOptions && !allowedOptions.has(optionValue)) {\n pushIssue(\n issues,\n `capabilities.${namespace}.fieldI18n.${fieldName}.optionLabelKeys.${optionValue}`,\n `option ${optionValue} is not defined in ${optionFieldName}`,\n )\n }\n }\n }\n }\n}\n\nfunction validateCapabilities(modelType, capabilities) {\n const issues = []\n if (!MODEL_TYPES.has(modelType)) {\n pushIssue(issues, 'type', 'type must be llm/image/video/audio/lipsync')\n return issues\n }\n if (capabilities === undefined || capabilities === null) return issues\n if (!isRecord(capabilities)) {\n pushIssue(issues, 'capabilities', 'capabilities must be an object')\n return issues\n }\n\n for (const namespace of Object.keys(capabilities)) {\n if (!CAPABILITY_NAMESPACES.has(namespace)) {\n pushIssue(issues, `capabilities.${namespace}`, `unknown capabilities namespace: ${namespace}`)\n continue\n }\n if (namespace !== modelType) {\n pushIssue(issues, `capabilities.${namespace}`, `namespace ${namespace} is not allowed for model type ${modelType}`)\n }\n }\n\n const llm = capabilities.llm\n if (llm !== undefined) {\n if (!isRecord(llm)) {\n pushIssue(issues, 'capabilities.llm', 'llm capabilities must be an object')\n } else {\n validateAllowedFields(issues, 'llm', llm)\n if (llm.reasoningEffortOptions !== undefined && !isStringArray(llm.reasoningEffortOptions)) {\n pushIssue(issues, 'capabilities.llm.reasoningEffortOptions', 'must be string array')\n }\n validateFieldI18nMap(issues, 'llm', llm)\n }\n }\n\n const image = capabilities.image\n if (image !== undefined) {\n if (!isRecord(image)) {\n pushIssue(issues, 'capabilities.image', 'image capabilities must be an object')\n } else {\n validateAllowedFields(issues, 'image', image)\n if (image.resolutionOptions !== undefined && !isStringArray(image.resolutionOptions)) {\n pushIssue(issues, 'capabilities.image.resolutionOptions', 'must be string array')\n }\n validateFieldI18nMap(issues, 'image', image)\n }\n }\n\n const video = capabilities.video\n if (video !== undefined) {\n if (!isRecord(video)) {\n pushIssue(issues, 'capabilities.video', 'video capabilities must be an object')\n } else {\n validateAllowedFields(issues, 'video', video)\n if (video.durationOptions !== undefined && !isNumberArray(video.durationOptions)) {\n pushIssue(issues, 'capabilities.video.durationOptions', 'must be number array')\n }\n if (video.fpsOptions !== undefined && !isNumberArray(video.fpsOptions)) {\n pushIssue(issues, 'capabilities.video.fpsOptions', 'must be number array')\n }\n if (video.resolutionOptions !== undefined && !isStringArray(video.resolutionOptions)) {\n pushIssue(issues, 'capabilities.video.resolutionOptions', 'must be string array')\n }\n if (video.supportGenerateAudio !== undefined && typeof video.supportGenerateAudio !== 'boolean') {\n pushIssue(issues, 'capabilities.video.supportGenerateAudio', 'must be boolean')\n }\n if (video.firstlastframe !== undefined && typeof video.firstlastframe !== 'boolean') {\n pushIssue(issues, 'capabilities.video.firstlastframe', 'must be boolean')\n }\n validateFieldI18nMap(issues, 'video', video)\n }\n }\n\n const audio = capabilities.audio\n if (audio !== undefined) {\n if (!isRecord(audio)) {\n pushIssue(issues, 'capabilities.audio', 'audio capabilities must be an object')\n } else {\n validateAllowedFields(issues, 'audio', audio)\n if (audio.voiceOptions !== undefined && !isStringArray(audio.voiceOptions)) {\n pushIssue(issues, 'capabilities.audio.voiceOptions', 'must be string array')\n }\n if (audio.rateOptions !== undefined && !isStringArray(audio.rateOptions)) {\n pushIssue(issues, 'capabilities.audio.rateOptions', 'must be string array')\n }\n validateFieldI18nMap(issues, 'audio', audio)\n }\n }\n\n const lipsync = capabilities.lipsync\n if (lipsync !== undefined) {\n if (!isRecord(lipsync)) {\n pushIssue(issues, 'capabilities.lipsync', 'lipsync capabilities must be an object')\n } else {\n validateAllowedFields(issues, 'lipsync', lipsync)\n if (lipsync.modeOptions !== undefined && !isStringArray(lipsync.modeOptions)) {\n pushIssue(issues, 'capabilities.lipsync.modeOptions', 'must be string array')\n }\n validateFieldI18nMap(issues, 'lipsync', lipsync)\n }\n }\n\n return issues\n}\n\nasync function main() {\n let PrismaClient\n try {\n ({ PrismaClient } = await import('@prisma/client'))\n } catch {\n throw new Error('MISSING_DEPENDENCY: @prisma/client is not installed, run npm install first')\n }\n\n prisma = new PrismaClient()\n const summary = {\n generatedAt: new Date().toISOString(),\n userPreference: {\n total: 0,\n invalidModelKeyFields: 0,\n invalidCustomModelsJson: 0,\n invalidCustomModelShape: 0,\n invalidCapabilities: 0,\n },\n novelPromotionProject: {\n total: 0,\n invalidModelKeyFields: 0,\n },\n samples: [],\n }\n\n const userPrefs = await prisma.userPreference.findMany({\n select: {\n id: true,\n customModels: true,\n analysisModel: true,\n characterModel: true,\n locationModel: true,\n storyboardModel: true,\n editModel: true,\n videoModel: true,\n },\n })\n\n for (const pref of userPrefs) {\n summary.userPreference.total += 1\n for (const field of MODEL_FIELDS) {\n const rawValue = pref[field]\n if (!rawValue) continue\n if (!parseModelKeyStrict(rawValue)) {\n summary.userPreference.invalidModelKeyFields += 1\n addSample(summary, {\n table: 'userPreference',\n rowId: pref.id,\n field,\n reason: 'model field is not provider::modelId',\n })\n }\n }\n\n if (!pref.customModels) continue\n let parsedCustomModels\n try {\n parsedCustomModels = JSON.parse(pref.customModels)\n } catch {\n summary.userPreference.invalidCustomModelsJson += 1\n addSample(summary, {\n table: 'userPreference',\n rowId: pref.id,\n field: 'customModels',\n reason: 'invalid JSON',\n })\n continue\n }\n if (!Array.isArray(parsedCustomModels)) {\n summary.userPreference.invalidCustomModelsJson += 1\n addSample(summary, {\n table: 'userPreference',\n rowId: pref.id,\n field: 'customModels',\n reason: 'customModels is not array',\n })\n continue\n }\n\n for (let index = 0; index < parsedCustomModels.length; index += 1) {\n const modelRaw = parsedCustomModels[index]\n if (!isRecord(modelRaw)) {\n summary.userPreference.invalidCustomModelShape += 1\n addSample(summary, {\n table: 'userPreference',\n rowId: pref.id,\n field: `customModels[${index}]`,\n reason: 'model item is not object',\n })\n continue\n }\n\n const modelKey = isNonEmptyString(modelRaw.modelKey) ? modelRaw.modelKey.trim() : ''\n const provider = isNonEmptyString(modelRaw.provider) ? modelRaw.provider.trim() : ''\n const modelId = isNonEmptyString(modelRaw.modelId) ? modelRaw.modelId.trim() : ''\n const parsed = parseModelKeyStrict(modelKey)\n if (!parsed || parsed.provider !== provider || parsed.modelId !== modelId) {\n summary.userPreference.invalidCustomModelShape += 1\n addSample(summary, {\n table: 'userPreference',\n rowId: pref.id,\n field: `customModels[${index}].modelKey`,\n reason: 'modelKey/provider/modelId mismatch',\n })\n }\n\n const modelType = isNonEmptyString(modelRaw.type) ? modelRaw.type.trim() : ''\n const capabilityIssues = validateCapabilities(modelType, modelRaw.capabilities)\n if (capabilityIssues.length > 0) {\n summary.userPreference.invalidCapabilities += 1\n addSample(summary, {\n table: 'userPreference',\n rowId: pref.id,\n field: capabilityIssues[0].field,\n reason: capabilityIssues[0].message,\n })\n }\n }\n }\n\n const projects = await prisma.novelPromotionProject.findMany({\n select: {\n id: true,\n analysisModel: true,\n characterModel: true,\n locationModel: true,\n storyboardModel: true,\n editModel: true,\n videoModel: true,\n },\n })\n\n for (const project of projects) {\n summary.novelPromotionProject.total += 1\n for (const field of MODEL_FIELDS) {\n const rawValue = project[field]\n if (!rawValue) continue\n if (!parseModelKeyStrict(rawValue)) {\n summary.novelPromotionProject.invalidModelKeyFields += 1\n addSample(summary, {\n table: 'novelPromotionProject',\n rowId: project.id,\n field,\n reason: 'model field is not provider::modelId',\n })\n }\n }\n }\n\n process.stdout.write(`${JSON.stringify(summary, null, 2)}\\n`)\n\n if (!STRICT) return\n const hasViolations = summary.userPreference.invalidModelKeyFields > 0\n || summary.userPreference.invalidCustomModelsJson > 0\n || summary.userPreference.invalidCustomModelShape > 0\n || summary.userPreference.invalidCapabilities > 0\n || summary.novelPromotionProject.invalidModelKeyFields > 0\n\n if (hasViolations) {\n process.exitCode = 1\n }\n}\n\nmain()\n .catch((error) => {\n process.stderr.write(`[check-model-config-contract] failed: ${String(error)}\\n`)\n process.exitCode = 1\n })\n .finally(async () => {\n if (prisma) {\n await prisma.$disconnect()\n }\n })\n"
},
{
"path": "scripts/check-no-console.ts",
"content": "import { execSync } from 'node:child_process'\n\nconst ALLOWLIST = new Set([\n 'src/lib/logging/core.ts',\n 'src/lib/logging/config.ts',\n 'src/lib/logging/context.ts',\n 'src/lib/logging/redact.ts',\n 'scripts/check-no-console.ts',\n 'scripts/guards/no-api-direct-llm-call.mjs',\n 'scripts/guards/no-internal-task-sync-fallback.mjs',\n 'scripts/guards/no-media-provider-bypass.mjs',\n 'scripts/guards/no-server-mirror-state.mjs',\n 'scripts/guards/task-loading-guard.mjs',\n 'scripts/guards/task-target-states-no-polling-guard.mjs',\n])\n\nfunction run(cmd: string): string {\n try {\n return execSync(cmd, { encoding: 'utf8' })\n } catch (error: unknown) {\n if (error && typeof error === 'object' && 'stdout' in error) {\n const stdout = (error as { stdout?: unknown }).stdout\n return typeof stdout === 'string' ? stdout : ''\n }\n return ''\n }\n}\n\nfunction main() {\n const output = run(`rg -n \"console\\\\\\\\.(log|info|warn|error|debug)\\\\\\\\(\" src scripts`)\n const lines = output\n .split('\\n')\n .map((line) => line.trim())\n .filter(Boolean)\n\n const violations = lines.filter((line) => {\n const file = line.split(':', 1)[0]\n return !ALLOWLIST.has(file)\n })\n\n if (violations.length > 0) {\n process.stderr.write('[check:logs] found forbidden console usage:\\n')\n for (const line of violations) {\n process.stderr.write(`- ${line}\\n`)\n }\n process.exit(1)\n }\n\n process.stdout.write(`[check:logs] ok scanned=${lines.length} allowlist=${ALLOWLIST.size}\\n`)\n}\n\nmain()\n"
},
{
"path": "scripts/check-outbound-image-runtime-sample.ts",
"content": "import { prisma } from '@/lib/prisma'\nimport { TASK_TYPE } from '@/lib/task/types'\n\ntype AnyJson = unknown\n\ntype Match = {\n path: string\n value: string\n}\n\ntype Options = {\n minutes: number\n limit: number\n projectId: string | null\n strictNoData: boolean\n includeEvents: boolean\n maxEventsPerTask: number\n json: boolean\n}\n\ntype FailureType = 'normalize' | 'model' | 'cancelled' | 'other'\n\nconst MODEL_ERROR_CODES = new Set([\n 'GENERATION_FAILED',\n 'GENERATION_TIMEOUT',\n 'RATE_LIMIT',\n 'EXTERNAL_ERROR',\n 'SENSITIVE_CONTENT',\n])\n\nfunction parseNumberArg(name: string, fallback: number): number {\n const raw = process.argv.find((arg) => arg.startsWith(`--${name}=`))\n if (!raw) return fallback\n const value = Number.parseInt(raw.split('=')[1] || '', 10)\n return Number.isFinite(value) && value > 0 ? value : fallback\n}\n\nfunction parseStringArg(name: string): string | null {\n const raw = process.argv.find((arg) => arg.startsWith(`--${name}=`))\n if (!raw) return null\n const value = (raw.split('=')[1] || '').trim()\n return value || null\n}\n\nfunction parseBooleanArg(name: string, fallback = false): boolean {\n const raw = process.argv.find((arg) => arg.startsWith(`--${name}=`))\n if (!raw) return fallback\n const value = (raw.split('=')[1] || '').trim().toLowerCase()\n return value === '1' || value === 'true' || value === 'yes' || value === 'on'\n}\n\nfunction parseOptions(): Options {\n return {\n minutes: parseNumberArg('minutes', 60 * 24),\n limit: parseNumberArg('limit', 200),\n projectId: parseStringArg('projectId'),\n strictNoData: parseBooleanArg('strictNoData', false),\n includeEvents: parseBooleanArg('includeEvents', false),\n maxEventsPerTask: parseNumberArg('maxEventsPerTask', 40),\n json: parseBooleanArg('json', false),\n }\n}\n\nfunction toExcerpt(value: string, max = 180): string {\n if (value.length <= max) return value\n return `${value.slice(0, max)}...`\n}\n\nfunction findStringMatches(\n value: AnyJson,\n predicate: (input: string) => boolean,\n path = '$',\n matches: Match[] = [],\n): Match[] {\n if (typeof value === 'string') {\n if (predicate(value)) matches.push({ path, value })\n return matches\n }\n if (Array.isArray(value)) {\n value.forEach((item, index) => {\n findStringMatches(item, predicate, `${path}[${index}]`, matches)\n })\n return matches\n }\n if (value && typeof value === 'object') {\n for (const [key, next] of Object.entries(value as Record)) {\n findStringMatches(next, predicate, `${path}.${key}`, matches)\n }\n }\n return matches\n}\n\nfunction classifyFailure(task: {\n errorCode: string | null\n errorMessage: string | null\n result: AnyJson | null\n events: Array<{ payload: AnyJson | null }>\n}): FailureType {\n const code = (task.errorCode || '').trim().toUpperCase()\n const normalizeRe = /normalize|video_frame_normalize|normalizeReferenceImagesForGeneration|reference image normalize failed|outbound image input is empty|relative_path_rejected/i\n const modelRe = /generation failed|provider|upstream|rate limit|timed out|timeout|sensitive/i\n\n if (code === 'TASK_CANCELLED') return 'cancelled'\n if (MODEL_ERROR_CODES.has(code)) return 'model'\n if (code) {\n const explicitNormalizeCode = code === 'INVALID_PARAMS' || code === 'OUTBOUND_IMAGE_FETCH_FAILED'\n if (explicitNormalizeCode) return 'normalize'\n return 'other'\n }\n\n const values: string[] = []\n if (code) values.push(code)\n if (task.errorMessage) values.push(task.errorMessage)\n if (task.result) {\n for (const hit of findStringMatches(task.result, () => true)) {\n values.push(hit.value)\n }\n }\n for (const event of task.events) {\n if (!event.payload) continue\n for (const hit of findStringMatches(event.payload, () => true)) {\n values.push(hit.value)\n }\n }\n\n if (values.some((item) => normalizeRe.test(item))) return 'normalize'\n if (values.some((item) => modelRe.test(item))) return 'model'\n return 'other'\n}\n\nasync function main() {\n const options = parseOptions()\n const since = new Date(Date.now() - options.minutes * 60_000)\n const monitoredTypes = [\n TASK_TYPE.MODIFY_ASSET_IMAGE,\n TASK_TYPE.ASSET_HUB_MODIFY,\n TASK_TYPE.VIDEO_PANEL,\n ]\n\n const tasks = await prisma.task.findMany({\n where: {\n type: { in: monitoredTypes },\n createdAt: { gte: since },\n ...(options.projectId ? { projectId: options.projectId } : {}),\n },\n select: {\n id: true,\n type: true,\n status: true,\n projectId: true,\n targetType: true,\n targetId: true,\n createdAt: true,\n errorCode: true,\n errorMessage: true,\n payload: true,\n result: true,\n },\n orderBy: { createdAt: 'desc' },\n take: options.limit,\n })\n\n if (tasks.length === 0) {\n process.stdout.write(\n `[check:outbound-image-runtime-sample] no data window=${options.minutes}m limit=${options.limit} strictNoData=${options.strictNoData}\\n`,\n )\n if (options.strictNoData) process.exit(2)\n return\n }\n\n const eventsByTaskId = new Map>()\n let eventCount = 0\n if (options.includeEvents) {\n for (const task of tasks) {\n const rows = await prisma.taskEvent.findMany({\n where: { taskId: task.id },\n select: {\n taskId: true,\n eventType: true,\n payload: true,\n createdAt: true,\n },\n orderBy: { id: 'desc' },\n take: options.maxEventsPerTask,\n })\n const ordered = [...rows].reverse()\n eventCount += ordered.length\n if (ordered.length > 0) {\n eventsByTaskId.set(\n task.id,\n ordered.map((event) => ({\n eventType: event.eventType,\n payload: event.payload,\n createdAt: event.createdAt,\n })),\n )\n }\n }\n }\n\n const nextImagePredicate = (input: string) => input.includes('/_next/image')\n const hits: Array<{\n taskId: string\n taskType: string\n source: 'task.payload' | 'task.result' | 'task.event'\n path: string\n value: string\n }> = []\n\n let failedCount = 0\n const failedByClass: Record = {\n normalize: 0,\n model: 0,\n cancelled: 0,\n other: 0,\n }\n const failedByCode: Record = {}\n\n for (const task of tasks) {\n const taskEventsForTask = eventsByTaskId.get(task.id) || []\n\n if (task.payload) {\n for (const match of findStringMatches(task.payload, nextImagePredicate)) {\n hits.push({\n taskId: task.id,\n taskType: task.type,\n source: 'task.payload',\n path: match.path,\n value: match.value,\n })\n }\n }\n\n if (task.result) {\n for (const match of findStringMatches(task.result, nextImagePredicate)) {\n hits.push({\n taskId: task.id,\n taskType: task.type,\n source: 'task.result',\n path: match.path,\n value: match.value,\n })\n }\n }\n\n for (const event of taskEventsForTask) {\n if (!event.payload) continue\n for (const match of findStringMatches(event.payload, nextImagePredicate)) {\n hits.push({\n taskId: task.id,\n taskType: task.type,\n source: 'task.event',\n path: match.path,\n value: match.value,\n })\n }\n }\n\n if (task.status === 'failed') {\n failedCount += 1\n const code = (task.errorCode || 'UNKNOWN').trim() || 'UNKNOWN'\n failedByCode[code] = (failedByCode[code] || 0) + 1\n const failureType = classifyFailure({\n errorCode: task.errorCode,\n errorMessage: task.errorMessage,\n result: task.result,\n events: taskEventsForTask,\n })\n failedByClass[failureType] += 1\n }\n }\n\n const typeCount = tasks.reduce>((acc, item) => {\n acc[item.type] = (acc[item.type] || 0) + 1\n return acc\n }, {})\n\n process.stdout.write(\n `[check:outbound-image-runtime-sample] window=${options.minutes}m sampled=${tasks.length} events=${eventCount} includeEvents=${options.includeEvents} next_image_hits=${hits.length}\\n`,\n )\n process.stdout.write(`[check:outbound-image-runtime-sample] task_types=${JSON.stringify(typeCount)}\\n`)\n process.stdout.write(\n `[check:outbound-image-runtime-sample] failures total=${failedCount} normalize=${failedByClass.normalize} model=${failedByClass.model} cancelled=${failedByClass.cancelled} other=${failedByClass.other} by_code=${JSON.stringify(failedByCode)}\\n`,\n )\n\n if (options.json) {\n process.stdout.write(\n `${JSON.stringify({\n windowMinutes: options.minutes,\n sampled: tasks.length,\n events: eventCount,\n includeEvents: options.includeEvents,\n nextImageHits: hits.length,\n taskTypes: typeCount,\n failures: {\n total: failedCount,\n byClass: failedByClass,\n byCode: failedByCode,\n },\n })}\\n`,\n )\n }\n\n if (hits.length > 0) {\n process.stderr.write('[check:outbound-image-runtime-sample] found /_next/image contamination:\\n')\n for (const hit of hits.slice(0, 20)) {\n process.stderr.write(\n `- task=${hit.taskId} type=${hit.taskType} source=${hit.source} path=${hit.path} value=${toExcerpt(hit.value)}\\n`,\n )\n }\n process.exit(1)\n }\n}\n\nmain()\n .catch((error) => {\n const message = error instanceof Error ? error.message : String(error)\n process.stderr.write(`[check:outbound-image-runtime-sample] failed: ${message}\\n`)\n process.exit(1)\n })\n .finally(async () => {\n await prisma.$disconnect()\n })\n"
},
{
"path": "scripts/check-outbound-image-success-rate.ts",
"content": "import { prisma } from '@/lib/prisma'\nimport { TASK_STATUS, TASK_TYPE } from '@/lib/task/types'\n\ntype StatusCount = Record\n\ntype WindowSummary = {\n total: number\n finishedTotal: number\n completed: number\n failed: number\n successRate: number | null\n byStatus: StatusCount\n byType: Record\n}\n\ntype Options = {\n minutes: number\n baselineMinutes: number\n baselineOffsetMinutes: number\n projectId: string | null\n tolerancePct: number\n minFinishedSamples: number\n strict: boolean\n json: boolean\n}\n\nconst DEFAULT_MINUTES = 60 * 24 * 7\nconst DEFAULT_TOLERANCE_PCT = 2\nconst DEFAULT_MIN_FINISHED_SAMPLES = 20\n\nfunction parseNumberArg(name: string, fallback: number): number {\n const raw = process.argv.find((arg) => arg.startsWith(`--${name}=`))\n if (!raw) return fallback\n const value = Number.parseFloat(raw.split('=')[1] || '')\n return Number.isFinite(value) && value > 0 ? value : fallback\n}\n\nfunction parseBooleanArg(name: string, fallback = false): boolean {\n const raw = process.argv.find((arg) => arg.startsWith(`--${name}=`))\n if (!raw) return fallback\n const value = (raw.split('=')[1] || '').trim().toLowerCase()\n return value === '1' || value === 'true' || value === 'yes' || value === 'on'\n}\n\nfunction parseStringArg(name: string): string | null {\n const raw = process.argv.find((arg) => arg.startsWith(`--${name}=`))\n if (!raw) return null\n const value = (raw.split('=')[1] || '').trim()\n return value || null\n}\n\nfunction parseOptions(): Options {\n const minutes = parseNumberArg('minutes', DEFAULT_MINUTES)\n const baselineMinutes = parseNumberArg('baselineMinutes', minutes)\n const baselineOffsetMinutes = parseNumberArg('baselineOffsetMinutes', minutes)\n return {\n minutes,\n baselineMinutes,\n baselineOffsetMinutes,\n projectId: parseStringArg('projectId'),\n tolerancePct: parseNumberArg('tolerancePct', DEFAULT_TOLERANCE_PCT),\n minFinishedSamples: parseNumberArg('minFinishedSamples', DEFAULT_MIN_FINISHED_SAMPLES),\n strict: parseBooleanArg('strict', false),\n json: parseBooleanArg('json', false),\n }\n}\n\nfunction asPct(value: number | null): string {\n return value === null ? 'N/A' : `${value.toFixed(2)}%`\n}\n\nfunction getSuccessRate(completed: number, failed: number): number | null {\n const total = completed + failed\n if (total <= 0) return null\n return (completed / total) * 100\n}\n\nfunction summarizeRows(\n rows: Array<{ status: string; type: string }>,\n): WindowSummary {\n const byStatus: StatusCount = {}\n const byType: Record = {}\n for (const row of rows) {\n byStatus[row.status] = (byStatus[row.status] || 0) + 1\n byType[row.type] = (byType[row.type] || 0) + 1\n }\n\n const completed = byStatus[TASK_STATUS.COMPLETED] || 0\n const failed = byStatus[TASK_STATUS.FAILED] || 0\n const finishedTotal = completed + failed\n\n return {\n total: rows.length,\n finishedTotal,\n completed,\n failed,\n successRate: getSuccessRate(completed, failed),\n byStatus,\n byType,\n }\n}\n\nasync function fetchWindowSummary(params: {\n from: Date\n to: Date\n projectId: string | null\n}) {\n const monitoredTypes = [\n TASK_TYPE.MODIFY_ASSET_IMAGE,\n TASK_TYPE.ASSET_HUB_MODIFY,\n TASK_TYPE.VIDEO_PANEL,\n ]\n\n const rows = await prisma.task.findMany({\n where: {\n type: { in: monitoredTypes },\n createdAt: {\n gte: params.from,\n lt: params.to,\n },\n ...(params.projectId ? { projectId: params.projectId } : {}),\n },\n select: {\n status: true,\n type: true,\n },\n })\n\n return summarizeRows(rows)\n}\n\nasync function main() {\n const options = parseOptions()\n const now = Date.now()\n\n const currentEnd = new Date(now)\n const currentStart = new Date(now - options.minutes * 60_000)\n\n const baselineEnd = new Date(now - options.baselineOffsetMinutes * 60_000)\n const baselineStart = new Date(baselineEnd.getTime() - options.baselineMinutes * 60_000)\n\n const [current, baseline] = await Promise.all([\n fetchWindowSummary({\n from: currentStart,\n to: currentEnd,\n projectId: options.projectId,\n }),\n fetchWindowSummary({\n from: baselineStart,\n to: baselineEnd,\n projectId: options.projectId,\n }),\n ])\n\n const hasEnoughCurrent = current.finishedTotal >= options.minFinishedSamples\n const hasEnoughBaseline = baseline.finishedTotal >= options.minFinishedSamples\n const hasEnoughSamples = hasEnoughCurrent && hasEnoughBaseline\n\n const rateDeltaPct =\n current.successRate !== null && baseline.successRate !== null\n ? current.successRate - baseline.successRate\n : null\n\n const meetsTolerance =\n rateDeltaPct !== null\n ? rateDeltaPct >= -Math.abs(options.tolerancePct)\n : false\n\n const status = hasEnoughSamples\n ? meetsTolerance\n ? 'pass'\n : 'fail'\n : 'blocked'\n\n process.stdout.write(\n `[check:outbound-image-success-rate] current=${asPct(current.successRate)} baseline=${asPct(baseline.successRate)} delta=${asPct(rateDeltaPct)} tolerance=-${Math.abs(options.tolerancePct).toFixed(2)}% status=${status}\\n`,\n )\n process.stdout.write(\n `[check:outbound-image-success-rate] current_finished=${current.finishedTotal} baseline_finished=${baseline.finishedTotal} min_required=${options.minFinishedSamples}\\n`,\n )\n process.stdout.write(\n `[check:outbound-image-success-rate] current_by_type=${JSON.stringify(current.byType)} baseline_by_type=${JSON.stringify(baseline.byType)}\\n`,\n )\n\n if (options.json) {\n process.stdout.write(\n `${JSON.stringify({\n status,\n tolerancePct: options.tolerancePct,\n minFinishedSamples: options.minFinishedSamples,\n windows: {\n current: {\n from: currentStart.toISOString(),\n to: currentEnd.toISOString(),\n ...current,\n },\n baseline: {\n from: baselineStart.toISOString(),\n to: baselineEnd.toISOString(),\n ...baseline,\n },\n },\n rateDeltaPct,\n hasEnoughSamples,\n })}\\n`,\n )\n }\n\n if (!options.strict) return\n\n if (status === 'pass') return\n if (status === 'blocked') process.exit(2)\n process.exit(1)\n}\n\nmain()\n .catch((error) => {\n const message = error instanceof Error ? error.message : String(error)\n process.stderr.write(`[check:outbound-image-success-rate] failed: ${message}\\n`)\n process.exit(1)\n })\n .finally(async () => {\n await prisma.$disconnect()\n })\n"
},
{
"path": "scripts/check-outbound-image-unification.ts",
"content": "import fs from 'node:fs'\nimport path from 'node:path'\n\ntype Rule = {\n file: string\n pattern: RegExp\n message: string\n}\n\nfunction readFile(relativePath: string): string {\n const fullPath = path.resolve(process.cwd(), relativePath)\n return fs.readFileSync(fullPath, 'utf8')\n}\n\nconst mustIncludeRules: Rule[] = [\n {\n file: 'src/lib/media/outbound-image.ts',\n pattern: /export\\s+async\\s+function\\s+normalizeToOriginalMediaUrl\\s*\\(/,\n message: 'missing normalizeToOriginalMediaUrl export',\n },\n {\n file: 'src/lib/media/outbound-image.ts',\n pattern: /export\\s+async\\s+function\\s+normalizeToBase64ForGeneration\\s*\\(/,\n message: 'missing normalizeToBase64ForGeneration export',\n },\n {\n file: 'src/lib/media/outbound-image.ts',\n pattern: /export\\s+async\\s+function\\s+normalizeReferenceImagesForGeneration\\s*\\(/,\n message: 'missing normalizeReferenceImagesForGeneration export',\n },\n {\n file: 'src/lib/media/outbound-image.ts',\n pattern: /class\\s+OutboundImageNormalizeError\\s+extends\\s+Error/,\n message: 'outbound-image.ts must expose structured normalize error type',\n },\n {\n file: 'src/lib/media/outbound-image.ts',\n pattern: /OUTBOUND_IMAGE_FETCH_FAILED/,\n message: 'outbound-image.ts must classify fetch failures with structured error codes',\n },\n {\n file: 'src/lib/media/outbound-image.ts',\n pattern: /OUTBOUND_IMAGE_REFERENCE_ALL_FAILED/,\n message: 'outbound-image.ts must fail explicitly when all references fail to normalize',\n },\n {\n file: 'src/lib/workers/handlers/image-task-handlers-core.ts',\n pattern: /normalizeToBase64ForGeneration\\(currentUrl\\)/,\n message: 'image-task-handlers-core.ts must convert currentUrl to base64 before outbound',\n },\n {\n file: 'src/lib/workers/handlers/image-task-handlers-core.ts',\n pattern: /normalizeReferenceImagesForGeneration\\(extraReferenceInputs\\)/,\n message: 'image-task-handlers-core.ts must normalize extra references before outbound',\n },\n {\n file: 'src/lib/workers/video.worker.ts',\n pattern: /const\\s+sourceImageBase64\\s*=\\s*await\\s+normalizeToBase64ForGeneration\\(sourceImageUrl\\)/,\n message: 'video.worker.ts must normalize source frame to base64',\n },\n {\n file: 'src/lib/workers/video.worker.ts',\n pattern: /lastFrameImageBase64\\s*=\\s*await\\s+normalizeToBase64ForGeneration\\(lastFrameUrl\\)/,\n message: 'video.worker.ts must normalize last frame to base64',\n },\n {\n file: 'src/app/api/novel-promotion/[projectId]/modify-asset-image/route.ts',\n pattern: /sanitizeImageInputsForTaskPayload/,\n message: 'modify-asset-image route must sanitize image inputs',\n },\n {\n file: 'src/app/api/novel-promotion/[projectId]/modify-storyboard-image/route.ts',\n pattern: /sanitizeImageInputsForTaskPayload/,\n message: 'modify-storyboard-image route must sanitize image inputs',\n },\n {\n file: 'src/app/api/asset-hub/modify-image/route.ts',\n pattern: /sanitizeImageInputsForTaskPayload/,\n message: 'asset-hub modify-image route must sanitize image inputs',\n },\n {\n file: 'src/components/ui/ImagePreviewModal.tsx',\n pattern: /import\\s+\\{\\s*resolveOriginalImageUrl,\\s*toDisplayImageUrl\\s*\\}\\s+from\\s+'@\\/lib\\/media\\/image-url'/,\n message: 'ImagePreviewModal must use shared image-url helpers',\n },\n {\n file: 'src/lib/novel-promotion/stages/video-stage-runtime-core.tsx',\n pattern: /onPreviewImage=\\{setPreviewImage\\}/,\n message: 'Video stage runtime must wire preview callback to VideoPanelCard',\n },\n {\n file: 'src/app/[locale]/workspace/[projectId]/modes/novel-promotion/components/video/panel-card/types.ts',\n pattern: /onPreviewImage\\?:\\s*\\(imageUrl:\\s*string\\)\\s*=>\\s*void/,\n message: 'VideoPanelCard runtime props must expose onPreviewImage',\n },\n {\n file: 'src/app/[locale]/workspace/[projectId]/modes/novel-promotion/components/video/panel-card/VideoPanelCardHeader.tsx',\n pattern: /className=\"absolute left-1\\/2 top-1\\/2 z-10 h-16 w-16 -translate-x-1\\/2 -translate-y-1\\/2 rounded-full\"/,\n message: 'VideoPanelCard play trigger must be centered small button (preview/play separation)',\n },\n]\n\nconst mustNotIncludeRules: Rule[] = [\n {\n file: 'src/lib/workers/handlers/image-task-handlers-core.ts',\n pattern: /referenceImages:\\s*\\[currentUrl\\]/,\n message: 'image-task-handlers-core.ts must not pass raw currentUrl directly as outbound reference',\n },\n {\n file: 'src/lib/workers/video.worker.ts',\n pattern: /imageUrl:\\s*sourceImageUrl/,\n message: 'video.worker.ts must not pass raw sourceImageUrl to generator',\n },\n {\n file: 'src/lib/media/outbound-image.ts',\n pattern: /return\\s+await\\s+toFetchableAbsoluteUrl\\(mediaPath\\)/,\n message: 'outbound-image.ts must not silently fallback when /m route cannot resolve storage key',\n },\n {\n file: 'src/lib/media/outbound-image.ts',\n pattern: /export\\s+async\\s+function\\s+imageUrlToBase64\\s*\\(/,\n message: 'outbound-image.ts must not keep legacy imageUrlToBase64 alias after phase 2 migration',\n },\n {\n file: 'src/lib/media/outbound-image.ts',\n pattern: /return\\s+await\\s+toFetchableAbsoluteUrl\\(unwrappedInput\\)/,\n message: 'outbound-image.ts must not silently fallback unknown inputs to fetchable url',\n },\n]\n\nfunction main() {\n const errors: string[] = []\n const cache = new Map()\n\n const getContent = (file: string) => {\n if (!cache.has(file)) cache.set(file, readFile(file))\n return cache.get(file) as string\n }\n\n for (const rule of mustIncludeRules) {\n const content = getContent(rule.file)\n if (!rule.pattern.test(content)) {\n errors.push(`${rule.file}: ${rule.message}`)\n }\n }\n\n for (const rule of mustNotIncludeRules) {\n const content = getContent(rule.file)\n if (rule.pattern.test(content)) {\n errors.push(`${rule.file}: ${rule.message}`)\n }\n }\n\n if (errors.length > 0) {\n process.stderr.write('[check:outbound-image-unification] found violations:\\n')\n for (const error of errors) {\n process.stderr.write(`- ${error}\\n`)\n }\n process.exit(1)\n }\n\n process.stdout.write(\n `[check:outbound-image-unification] ok include_checks=${mustIncludeRules.length} exclude_checks=${mustNotIncludeRules.length}\\n`,\n )\n}\n\nmain()\n"
},
{
"path": "scripts/check-pricing-catalog.mjs",
"content": "import { promises as fs } from 'node:fs'\nimport path from 'node:path'\n\nconst CATALOG_DIR = path.resolve(process.cwd(), 'standards/pricing')\nconst CAPABILITY_CATALOG_FILE = path.resolve(process.cwd(), 'standards/capabilities/image-video.catalog.json')\nconst API_TYPES = new Set(['text', 'image', 'video', 'voice', 'voice-design', 'lip-sync'])\nconst PRICING_MODES = new Set(['flat', 'capability'])\nconst TEXT_TOKEN_TYPES = new Set(['input', 'output'])\n\nfunction isRecord(value) {\n return !!value && typeof value === 'object' && !Array.isArray(value)\n}\n\nfunction isNonEmptyString(value) {\n return typeof value === 'string' && value.trim().length > 0\n}\n\nfunction isCapabilityValue(value) {\n return typeof value === 'string' || typeof value === 'number' || typeof value === 'boolean'\n}\n\nfunction isFiniteNumber(value) {\n return typeof value === 'number' && Number.isFinite(value)\n}\n\nfunction pushIssue(issues, file, index, field, message) {\n issues.push({ file, index, field, message })\n}\n\nfunction getProviderKey(providerId) {\n const marker = providerId.indexOf(':')\n return marker === -1 ? providerId : providerId.slice(0, marker)\n}\n\nfunction buildModelKey(modelType, provider, modelId) {\n return `${modelType}::${provider}::${modelId}`\n}\n\nasync function listCatalogFiles() {\n const entries = await fs.readdir(CATALOG_DIR, { withFileTypes: true })\n return entries\n .filter((entry) => entry.isFile() && entry.name.endsWith('.json'))\n .map((entry) => path.join(CATALOG_DIR, entry.name))\n}\n\nasync function readCatalog(filePath) {\n const raw = await fs.readFile(filePath, 'utf8')\n const parsed = JSON.parse(raw)\n if (!Array.isArray(parsed)) {\n throw new Error(`catalog must be an array: ${filePath}`)\n }\n return parsed\n}\n\nasync function readCapabilityCatalog() {\n const raw = await fs.readFile(CAPABILITY_CATALOG_FILE, 'utf8')\n const parsed = JSON.parse(raw)\n if (!Array.isArray(parsed)) {\n throw new Error(`capability catalog must be an array: ${CAPABILITY_CATALOG_FILE}`)\n }\n return parsed\n}\n\nfunction extractCapabilityOptionFields(modelType, capabilities) {\n if (!isRecord(capabilities)) return new Set()\n const namespace = capabilities[modelType]\n if (!isRecord(namespace)) return new Set()\n\n const fields = new Set()\n for (const [key, value] of Object.entries(namespace)) {\n if (!key.endsWith('Options')) continue\n if (!Array.isArray(value) || value.length === 0) continue\n const field = key.slice(0, -'Options'.length)\n fields.add(field)\n }\n return fields\n}\n\nfunction buildCapabilityOptionFieldMap(capabilityEntries) {\n const map = new Map()\n for (const entry of capabilityEntries) {\n if (!isRecord(entry)) continue\n const modelType = typeof entry.modelType === 'string' ? entry.modelType.trim() : ''\n const provider = typeof entry.provider === 'string' ? entry.provider.trim() : ''\n const modelId = typeof entry.modelId === 'string' ? entry.modelId.trim() : ''\n if (!modelType || !provider || !modelId) continue\n\n const fields = extractCapabilityOptionFields(modelType, entry.capabilities)\n map.set(buildModelKey(modelType, provider, modelId), fields)\n const providerKey = getProviderKey(provider)\n const fallbackKey = buildModelKey(modelType, providerKey, modelId)\n if (!map.has(fallbackKey)) {\n map.set(fallbackKey, fields)\n }\n }\n return map\n}\n\nfunction validateTier(issues, file, index, tier, tierIndex) {\n if (!isRecord(tier)) {\n pushIssue(issues, file, index, `pricing.tiers[${tierIndex}]`, 'tier must be object')\n return\n }\n\n if (!isRecord(tier.when) || Object.keys(tier.when).length === 0) {\n pushIssue(issues, file, index, `pricing.tiers[${tierIndex}].when`, 'when must be non-empty object')\n } else {\n for (const [field, value] of Object.entries(tier.when)) {\n if (!isCapabilityValue(value)) {\n pushIssue(\n issues,\n file,\n index,\n `pricing.tiers[${tierIndex}].when.${field}`,\n 'condition value must be string/number/boolean',\n )\n }\n }\n }\n\n if (!isFiniteNumber(tier.amount) || tier.amount < 0) {\n pushIssue(issues, file, index, `pricing.tiers[${tierIndex}].amount`, 'amount must be finite number >= 0')\n }\n}\n\nfunction validateTextCapabilityTiers(issues, file, index, tiers) {\n const seenTokenTypes = new Set()\n\n for (let tierIndex = 0; tierIndex < tiers.length; tierIndex += 1) {\n const tier = tiers[tierIndex]\n if (!isRecord(tier) || !isRecord(tier.when)) continue\n\n const whenFields = Object.keys(tier.when)\n if (whenFields.length !== 1 || whenFields[0] !== 'tokenType') {\n pushIssue(issues, file, index, `pricing.tiers[${tierIndex}].when`, 'text capability tier must only contain tokenType')\n continue\n }\n\n const tokenType = tier.when.tokenType\n if (typeof tokenType !== 'string' || !TEXT_TOKEN_TYPES.has(tokenType)) {\n pushIssue(issues, file, index, `pricing.tiers[${tierIndex}].when.tokenType`, 'tokenType must be input or output')\n continue\n }\n\n if (seenTokenTypes.has(tokenType)) {\n pushIssue(issues, file, index, `pricing.tiers[${tierIndex}].when.tokenType`, `duplicate tokenType tier: ${tokenType}`)\n continue\n }\n seenTokenTypes.add(tokenType)\n }\n\n for (const requiredTokenType of TEXT_TOKEN_TYPES) {\n if (!seenTokenTypes.has(requiredTokenType)) {\n pushIssue(issues, file, index, 'pricing.tiers', `missing text tier tokenType=${requiredTokenType}`)\n }\n }\n}\n\nfunction validateMediaCapabilityTierFields(issues, file, index, item, tiers, capabilityOptionFieldsMap) {\n const modelType = item.apiType\n const provider = item.provider\n const modelId = item.modelId\n const modelKey = buildModelKey(modelType, provider, modelId)\n const fallbackKey = buildModelKey(modelType, getProviderKey(provider), modelId)\n const optionFields = capabilityOptionFieldsMap.get(modelKey) || capabilityOptionFieldsMap.get(fallbackKey)\n\n if (!optionFields || optionFields.size === 0) {\n pushIssue(issues, file, index, 'pricing.tiers', `no capability option fields found for ${modelType} ${provider}/${modelId}`)\n return\n }\n\n for (let tierIndex = 0; tierIndex < tiers.length; tierIndex += 1) {\n const tier = tiers[tierIndex]\n if (!isRecord(tier) || !isRecord(tier.when)) continue\n for (const field of Object.keys(tier.when)) {\n if (!optionFields.has(field)) {\n pushIssue(\n issues,\n file,\n index,\n `pricing.tiers[${tierIndex}].when.${field}`,\n `field ${field} is not declared in capabilities options for ${modelType} ${provider}/${modelId}`,\n )\n }\n }\n }\n}\n\nfunction validateDuplicateCapabilityTiers(issues, file, index, tiers) {\n const seen = new Set()\n for (let tierIndex = 0; tierIndex < tiers.length; tierIndex += 1) {\n const tier = tiers[tierIndex]\n if (!isRecord(tier) || !isRecord(tier.when)) continue\n const signature = JSON.stringify(Object.entries(tier.when).sort((left, right) => left[0].localeCompare(right[0])))\n if (seen.has(signature)) {\n pushIssue(issues, file, index, `pricing.tiers[${tierIndex}].when`, 'duplicate capability tier condition')\n continue\n }\n seen.add(signature)\n }\n}\n\nfunction validatePricing(issues, file, index, item, capabilityOptionFieldsMap) {\n const pricing = item.pricing\n if (!isRecord(pricing)) {\n pushIssue(issues, file, index, 'pricing', 'pricing must be object')\n return\n }\n\n if (!isNonEmptyString(pricing.mode) || !PRICING_MODES.has(pricing.mode)) {\n pushIssue(issues, file, index, 'pricing.mode', 'pricing.mode must be flat or capability')\n return\n }\n\n if (pricing.mode === 'flat') {\n if (!isFiniteNumber(pricing.flatAmount) || pricing.flatAmount < 0) {\n pushIssue(issues, file, index, 'pricing.flatAmount', 'flatAmount must be finite number >= 0')\n }\n return\n }\n\n if (!Array.isArray(pricing.tiers) || pricing.tiers.length === 0) {\n pushIssue(issues, file, index, 'pricing.tiers', 'tiers must be non-empty array')\n return\n }\n\n for (let tierIndex = 0; tierIndex < pricing.tiers.length; tierIndex += 1) {\n validateTier(issues, file, index, pricing.tiers[tierIndex], tierIndex)\n }\n\n validateDuplicateCapabilityTiers(issues, file, index, pricing.tiers)\n\n if (item.apiType === 'text') {\n validateTextCapabilityTiers(issues, file, index, pricing.tiers)\n return\n }\n\n if (item.apiType === 'image' || item.apiType === 'video') {\n validateMediaCapabilityTierFields(issues, file, index, item, pricing.tiers, capabilityOptionFieldsMap)\n }\n}\n\nasync function main() {\n const issues = []\n const files = await listCatalogFiles()\n const capabilityCatalog = await readCapabilityCatalog()\n const capabilityOptionFieldsMap = buildCapabilityOptionFieldMap(capabilityCatalog)\n if (files.length === 0) {\n throw new Error(`no pricing files found in ${CATALOG_DIR}`)\n }\n\n for (const filePath of files) {\n const items = await readCatalog(filePath)\n for (let index = 0; index < items.length; index += 1) {\n const item = items[index]\n if (!isRecord(item)) {\n pushIssue(issues, filePath, index, 'entry', 'entry must be object')\n continue\n }\n\n if (!isNonEmptyString(item.apiType) || !API_TYPES.has(item.apiType)) {\n pushIssue(issues, filePath, index, 'apiType', 'apiType must be one of text/image/video/voice/voice-design/lip-sync')\n }\n if (!isNonEmptyString(item.provider)) {\n pushIssue(issues, filePath, index, 'provider', 'provider must be non-empty string')\n }\n if (!isNonEmptyString(item.modelId)) {\n pushIssue(issues, filePath, index, 'modelId', 'modelId must be non-empty string')\n }\n\n validatePricing(issues, filePath, index, item, capabilityOptionFieldsMap)\n }\n }\n\n if (issues.length === 0) {\n process.stdout.write(`[check-pricing-catalog] OK (${files.length} files)\\n`)\n return\n }\n\n const maxPrint = 50\n for (const issue of issues.slice(0, maxPrint)) {\n process.stdout.write(`[check-pricing-catalog] ${issue.file}#${issue.index} ${issue.field}: ${issue.message}\\n`)\n }\n if (issues.length > maxPrint) {\n process.stdout.write(`[check-pricing-catalog] ... ${issues.length - maxPrint} more issues\\n`)\n }\n process.exitCode = 1\n}\n\nmain().catch((error) => {\n process.stderr.write(`[check-pricing-catalog] failed: ${String(error)}\\n`)\n process.exitCode = 1\n})\n"
},
{
"path": "scripts/cleanup-remove-legacy-voice-data.ts",
"content": "import { prisma } from '@/lib/prisma'\n\ntype CharacterVoiceRecord = {\n id: string\n customVoiceUrl: string | null\n}\n\ntype SpeakerVoiceConfig = {\n voiceType?: unknown\n voiceId?: unknown\n audioUrl?: unknown\n [key: string]: unknown\n}\n\ntype CleanupSummary = {\n projectCharactersUpdated: number\n globalCharactersUpdated: number\n episodeSpeakerVoicesUpdated: number\n episodeSpeakerVoicesCleared: number\n invalidSpeakerVoicesSkipped: number\n}\n\nfunction hasPlayableAudioUrl(value: unknown) {\n return typeof value === 'string' && value.trim().length > 0\n}\n\nfunction normalizeVoiceType(customVoiceUrl: string | null) {\n return hasPlayableAudioUrl(customVoiceUrl) ? 'custom' : null\n}\n\nasync function cleanupCharacterTable(records: CharacterVoiceRecord[], table: 'project' | 'global') {\n let updated = 0\n for (const row of records) {\n const nextVoiceType = normalizeVoiceType(row.customVoiceUrl)\n if (table === 'project') {\n await prisma.novelPromotionCharacter.update({\n where: { id: row.id },\n data: {\n voiceType: nextVoiceType,\n voiceId: null,\n },\n })\n } else {\n await prisma.globalCharacter.update({\n where: { id: row.id },\n data: {\n voiceType: nextVoiceType,\n voiceId: null,\n },\n })\n }\n updated += 1\n }\n return updated\n}\n\nfunction normalizeSpeakerVoices(payload: string): {\n ok: true\n changed: boolean\n cleared: boolean\n next: string | null\n} | {\n ok: false\n} {\n let parsed: unknown\n try {\n parsed = JSON.parse(payload)\n } catch {\n return { ok: false }\n }\n\n if (!parsed || typeof parsed !== 'object' || Array.isArray(parsed)) {\n return { ok: false }\n }\n\n const source = parsed as Record\n const next: Record = {}\n let changed = false\n\n for (const [speaker, value] of Object.entries(source)) {\n if (!value || typeof value !== 'object' || Array.isArray(value)) {\n return { ok: false }\n }\n\n const config = { ...(value as SpeakerVoiceConfig) }\n if (config.voiceType === 'azure') {\n if (hasPlayableAudioUrl(config.audioUrl)) {\n config.voiceType = 'custom'\n config.voiceId = null\n next[speaker] = config\n } else {\n // No usable audio, drop stale azure speaker config.\n }\n changed = true\n continue\n }\n\n next[speaker] = config\n }\n\n const keys = Object.keys(next)\n if (keys.length === 0) {\n return {\n ok: true,\n changed,\n cleared: true,\n next: null,\n }\n }\n\n return {\n ok: true,\n changed,\n cleared: false,\n next: changed ? JSON.stringify(next) : payload,\n }\n}\n\nasync function main() {\n const summary: CleanupSummary = {\n projectCharactersUpdated: 0,\n globalCharactersUpdated: 0,\n episodeSpeakerVoicesUpdated: 0,\n episodeSpeakerVoicesCleared: 0,\n invalidSpeakerVoicesSkipped: 0,\n }\n\n const [projectCharacters, globalCharacters] = await Promise.all([\n prisma.novelPromotionCharacter.findMany({\n where: { voiceType: 'azure' },\n select: {\n id: true,\n customVoiceUrl: true,\n },\n }),\n prisma.globalCharacter.findMany({\n where: { voiceType: 'azure' },\n select: {\n id: true,\n customVoiceUrl: true,\n },\n }),\n ])\n\n summary.projectCharactersUpdated = await cleanupCharacterTable(projectCharacters, 'project')\n summary.globalCharactersUpdated = await cleanupCharacterTable(globalCharacters, 'global')\n\n const episodes = await prisma.novelPromotionEpisode.findMany({\n where: {\n speakerVoices: { not: null },\n },\n select: {\n id: true,\n speakerVoices: true,\n },\n })\n\n for (const row of episodes) {\n const speakerVoices = row.speakerVoices\n if (!speakerVoices || !speakerVoices.includes('\"voiceType\":\"azure\"')) {\n continue\n }\n const normalized = normalizeSpeakerVoices(speakerVoices)\n if (!normalized.ok) {\n summary.invalidSpeakerVoicesSkipped += 1\n continue\n }\n if (!normalized.changed) {\n continue\n }\n await prisma.novelPromotionEpisode.update({\n where: { id: row.id },\n data: {\n speakerVoices: normalized.next,\n },\n })\n summary.episodeSpeakerVoicesUpdated += 1\n if (normalized.cleared) {\n summary.episodeSpeakerVoicesCleared += 1\n }\n }\n\n process.stdout.write(`${JSON.stringify({\n ok: true,\n checkedAt: new Date().toISOString(),\n summary,\n }, null, 2)}\\n`)\n}\n\nmain()\n .catch((error) => {\n process.stderr.write(`${error instanceof Error ? error.stack || error.message : String(error)}\\n`)\n process.exitCode = 1\n })\n .finally(async () => {\n await prisma.$disconnect()\n })\n"
},
{
"path": "scripts/diagnose-project.ts",
"content": "/**\n * 诊断项目任务状态\n * 运行: npx tsx scripts/diagnose-project.ts \n */\nimport { config } from 'dotenv'\nconfig()\n\nimport { prisma } from '../src/lib/prisma'\n\nasync function diagnoseProject(projectId: string) {\n console.log(`🔍 诊断项目: ${projectId}\\n`)\n\n // 1. 检查项目是否存在\n console.log('1️⃣ 项目基本信息:')\n const project = await prisma.project.findUnique({\n where: { id: projectId },\n include: {\n novelPromotionData: true\n }\n })\n \n if (!project) {\n console.log(' ❌ 项目不存在')\n process.exit(1)\n }\n \n console.log(` 名称: ${project.name}`)\n console.log(` 模式: ${project.mode}`)\n console.log(` 用户ID: ${project.userId}`)\n\n // 2. 检查 NovelPromotionProject\n console.log('\\n2️⃣ 小说推广项目配置:')\n const novelData = project.novelPromotionData\n if (!novelData) {\n console.log(' ❌ novelPromotionData 未创建')\n } else {\n console.log(` ID: ${novelData.id}`)\n console.log(` 视频比例: ${novelData.videoRatio || '未设置'}`)\n console.log(` 画风提示: ${novelData.artStylePrompt || '未设置'}`)\n }\n\n // 3. 检查场景和场景图片\n console.log('\\n3️⃣ 场景资产:')\n const novelProjectId = novelData?.id\n if (!novelProjectId) {\n console.log(' ❌ 无法获取 novelPromotionProject ID')\n process.exit(1)\n }\n \n const locations = await prisma.novelPromotionLocation.findMany({\n where: { novelPromotionProjectId: novelProjectId },\n include: {\n images: {\n orderBy: { imageIndex: 'asc' }\n }\n }\n })\n \n console.log(` 场景数量: ${locations.length}`)\n \n for (const loc of locations) {\n console.log(`\\n 📍 ${loc.name} (${loc.id})`)\n console.log(` 图片数量: ${loc.images?.length || 0}`)\n \n for (const img of loc.images || []) {\n console.log(` - [${img.imageIndex}] imageUrl: ${img.imageUrl || 'null'}`)\n console.log(` isSelected: ${img.isSelected}`)\n console.log(` description: ${img.description || 'null'}`)\n\n // 检查 MediaObject\n if (img.imageUrl) {\n const media = await prisma.mediaObject.findFirst({\n where: { \n OR: [\n { storageKey: img.imageUrl },\n { storageKey: { contains: img.imageUrl.split('/').pop() || '' } }\n ]\n }\n })\n if (media) {\n console.log(` ✅ MediaObject: ${media.publicId}`)\n } else {\n console.log(` ⚠️ 未找到 MediaObject`)\n }\n }\n }\n }\n\n // 4. 检查最近的任务\n console.log('\\n4️⃣ 最近的任务:')\n const tasks = await prisma.task.findMany({\n where: { projectId },\n orderBy: { createdAt: 'desc' },\n take: 10\n })\n \n console.log(` 任务数量: ${tasks.length}`)\n \n for (const task of tasks) {\n console.log(`\\n 📝 ${task.type} (${task.id})`)\n console.log(` 状态: ${task.status}`)\n console.log(` 目标: ${task.targetType} / ${task.targetId}`)\n console.log(` 创建时间: ${task.createdAt}`)\n console.log(` 更新时间: ${task.updatedAt}`)\n\n if (task.errorMessage || task.errorCode) {\n console.log(` ❌ 错误码: ${task.errorCode || 'N/A'}`)\n console.log(` ❌ 错误信息: ${task.errorMessage?.substring(0, 200) || 'N/A'}`)\n }\n\n // 获取任务事件\n const events = await prisma.taskEvent.findMany({\n where: { taskId: task.id },\n orderBy: { createdAt: 'desc' },\n take: 3\n })\n \n if (events.length > 0) {\n console.log(` 最近事件:`)\n for (const event of events) {\n console.log(` - ${event.eventType}: ${JSON.stringify(event.payload).substring(0, 100)}`)\n }\n }\n }\n\n // 5. 检查 Worker 队列状态\n console.log('\\n5️⃣ 检查 Worker 配置:')\n console.log(` REDIS_HOST: ${process.env.REDIS_HOST || '未设置'}`)\n console.log(` REDIS_PORT: ${process.env.REDIS_PORT || '未设置'}`)\n \n // 尝试连接 Redis\n try {\n const { Redis } = await import('ioredis')\n const redis = new Redis({\n host: process.env.REDIS_HOST || 'localhost',\n port: parseInt(process.env.REDIS_PORT || '6379'),\n maxRetriesPerRequest: 3,\n connectTimeout: 5000\n })\n \n const pingResult = await redis.ping()\n console.log(` ✅ Redis 连接: ${pingResult}`)\n \n // 检查 BullMQ 队列\n const queueKeys = await redis.keys('bull:*:id')\n console.log(` BullMQ 队列数量: ${queueKeys.length}`)\n \n for (const key of queueKeys.slice(0, 5)) {\n const queueName = key.replace('bull:', '').replace(':id', '')\n const jobCounts = await redis.hgetall(`bull:${queueName}:id`)\n console.log(` - ${queueName}`)\n }\n \n redis.disconnect()\n } catch (error) {\n console.log(` ❌ Redis 连接失败:`, error)\n }\n\n // 6. 检查模型配置\n console.log('\\n6️⃣ 检查用户模型配置:')\n const userPreference = await prisma.userPreference.findUnique({\n where: { userId: project.userId }\n })\n\n if (!userPreference) {\n console.log(' ❌ 用户偏好配置不存在')\n } else {\n console.log(` 角色模型: ${userPreference.characterModel || '未设置'}`)\n console.log(` 场景模型: ${userPreference.locationModel || '未设置'}`)\n console.log(` 视频模型: ${userPreference.videoModel || '未设置'}`)\n console.log(` 编辑模型: ${userPreference.editModel || '未设置'}`)\n console.log(` 口型同步模型: ${userPreference.lipSyncModel || '未设置'}`)\n console.log(` 分析模型: ${userPreference.analysisModel || '未设置'}`)\n }\n\n console.log('\\n✨ 诊断完成!')\n \n await prisma.$disconnect()\n}\n\nconst projectId = process.argv[2]\nif (!projectId) {\n console.log('用法: npx tsx scripts/diagnose-project.ts ')\n console.log('示例: npx tsx scripts/diagnose-project.ts fae709e9-9215-4b3f-9f53-dad871f09896')\n process.exit(1)\n}\n\ndiagnoseProject(projectId).catch(console.error)\n"
},
{
"path": "scripts/guards/api-route-contract-guard.mjs",
"content": "#!/usr/bin/env node\n\nimport fs from 'fs'\nimport path from 'path'\nimport process from 'process'\nimport { pathToFileURL } from 'url'\n\nconst root = process.cwd()\nconst apiDir = path.join(root, 'src', 'app', 'api')\n\nexport const API_HANDLER_ALLOWLIST = new Set([\n 'src/app/api/auth/[...nextauth]/route.ts',\n 'src/app/api/files/[...path]/route.ts',\n 'src/app/api/system/boot-id/route.ts',\n])\n\nexport const PUBLIC_ROUTE_ALLOWLIST = new Set([\n 'src/app/api/auth/[...nextauth]/route.ts',\n 'src/app/api/auth/register/route.ts',\n 'src/app/api/cos/image/route.ts',\n 'src/app/api/files/[...path]/route.ts',\n 'src/app/api/storage/sign/route.ts',\n 'src/app/api/system/boot-id/route.ts',\n])\n\nconst AUTH_CALL_PATTERNS = [\n /\\brequireUserAuth\\s*\\(/,\n /\\brequireProjectAuth\\s*\\(/,\n /\\brequireProjectAuthLight\\s*\\(/,\n]\n\nfunction fail(title, details = []) {\n process.stderr.write(`\\n[api-route-contract-guard] ${title}\\n`)\n for (const detail of details) {\n process.stderr.write(` - ${detail}\\n`)\n }\n process.exit(1)\n}\n\nfunction walk(dir, out = []) {\n if (!fs.existsSync(dir)) return out\n const entries = fs.readdirSync(dir, { withFileTypes: true })\n for (const entry of entries) {\n if (entry.name === '.git' || entry.name === '.next' || entry.name === 'node_modules') continue\n const fullPath = path.join(dir, entry.name)\n if (entry.isDirectory()) {\n walk(fullPath, out)\n continue\n }\n if (entry.name === 'route.ts') out.push(fullPath)\n }\n return out\n}\n\nfunction toRel(fullPath) {\n return path.relative(root, fullPath).split(path.sep).join('/')\n}\n\nfunction hasApiHandlerWrapper(content) {\n return /\\bapiHandler\\s*\\(/.test(content)\n}\n\nfunction hasRequiredAuth(content) {\n return AUTH_CALL_PATTERNS.some((pattern) => pattern.test(content))\n}\n\nexport function inspectRouteContract(relPath, content) {\n const violations = []\n\n if (!API_HANDLER_ALLOWLIST.has(relPath) && !hasApiHandlerWrapper(content)) {\n violations.push(`${relPath} missing apiHandler wrapper`)\n }\n\n if (!PUBLIC_ROUTE_ALLOWLIST.has(relPath) && !hasRequiredAuth(content)) {\n violations.push(`${relPath} missing requireUserAuth/requireProjectAuth/requireProjectAuthLight`)\n }\n\n return violations\n}\n\nexport function findApiRouteContractViolations(scanRoot = root) {\n const routesRoot = path.join(scanRoot, 'src', 'app', 'api')\n return walk(routesRoot)\n .map((fullPath) => {\n const relPath = path.relative(scanRoot, fullPath).split(path.sep).join('/')\n const content = fs.readFileSync(fullPath, 'utf8')\n return inspectRouteContract(relPath, content)\n })\n .flat()\n}\n\nexport function main() {\n if (!fs.existsSync(apiDir)) {\n fail('Missing src/app/api directory')\n }\n\n const violations = walk(apiDir)\n .map((fullPath) => {\n const relPath = toRel(fullPath)\n const content = fs.readFileSync(fullPath, 'utf8')\n return inspectRouteContract(relPath, content)\n })\n .flat()\n\n if (violations.length > 0) {\n fail('Found API route contract violations', violations)\n }\n\n process.stdout.write(\n `[api-route-contract-guard] OK routes=${walk(apiDir).length} public=${PUBLIC_ROUTE_ALLOWLIST.size} apiHandlerExceptions=${API_HANDLER_ALLOWLIST.size}\\n`,\n )\n}\n\nif (process.argv[1] && import.meta.url === pathToFileURL(process.argv[1]).href) {\n main()\n}\n"
},
{
"path": "scripts/guards/changed-file-test-impact-guard.mjs",
"content": "#!/usr/bin/env node\n\nimport { execSync } from 'node:child_process'\nimport { pathToFileURL } from 'node:url'\n\nconst RULES = [\n {\n name: 'api',\n source: /^src\\/app\\/api\\//,\n tests: [/^tests\\/integration\\/api\\/contract\\//, /^tests\\/system\\//, /^tests\\/regression\\//],\n message: 'changing src/app/api/** requires a matching contract, system, or regression test change',\n },\n {\n name: 'worker',\n source: /^src\\/lib\\/workers\\//,\n tests: [/^tests\\/unit\\/worker\\//, /^tests\\/system\\//, /^tests\\/regression\\//],\n message: 'changing src/lib/workers/** requires a matching worker, system, or regression test change',\n },\n {\n name: 'task',\n source: /^src\\/lib\\/task\\//,\n tests: [/^tests\\/unit\\/task\\//, /^tests\\/system\\//, /^tests\\/regression\\//],\n message: 'changing src/lib/task/** requires a matching task, system, or regression test change',\n },\n {\n name: 'media',\n source: /^src\\/lib\\/media\\//,\n tests: [/^tests\\/unit\\//, /^tests\\/system\\//, /^tests\\/regression\\//],\n message: 'changing src/lib/media/** requires a matching unit, system, or regression test change',\n },\n {\n name: 'provider',\n source: /^src\\/lib\\/(generator-api|generators|model-gateway|lipsync|providers)\\//,\n tests: [/^tests\\/unit\\/(providers|model-gateway|llm)\\//, /^tests\\/integration\\/provider\\//, /^tests\\/system\\//, /^tests\\/regression\\//],\n message: 'changing provider/gateway code requires provider contract, system, or regression test change',\n },\n]\n\nfunction normalizeChangedFiles(rawFiles) {\n return rawFiles\n .flatMap((item) => item.split(/[\\n,]/))\n .map((item) => item.trim())\n .filter(Boolean)\n}\n\nfunction readGitChangedFiles() {\n try {\n const output = execSync('git diff --name-only --cached', {\n cwd: process.cwd(),\n encoding: 'utf8',\n stdio: ['ignore', 'pipe', 'ignore'],\n })\n return normalizeChangedFiles([output])\n } catch {\n return []\n }\n}\n\nexport function inspectChangedFiles(changedFiles) {\n const changed = normalizeChangedFiles(changedFiles)\n const changedTests = changed.filter((file) => file.startsWith('tests/'))\n const violations = []\n\n for (const rule of RULES) {\n const impactedSources = changed.filter((file) => rule.source.test(file))\n if (impactedSources.length === 0) continue\n const hasMatchingTestChange = changedTests.some((file) => rule.tests.some((pattern) => pattern.test(file)))\n if (!hasMatchingTestChange) {\n violations.push(`${rule.name}: ${rule.message}; sources=${impactedSources.join(',')}`)\n }\n }\n\n return violations\n}\n\nfunction fail(violations) {\n console.error('\\n[changed-file-test-impact-guard] Missing matching test changes')\n for (const violation of violations) {\n console.error(` - ${violation}`)\n }\n process.exit(1)\n}\n\nfunction runCli() {\n const inputFiles = process.argv.slice(2)\n const changedFiles = inputFiles.length > 0\n ? normalizeChangedFiles(inputFiles)\n : normalizeChangedFiles([process.env.TEST_IMPACT_CHANGED_FILES || '', ...readGitChangedFiles()])\n\n if (changedFiles.length === 0) {\n console.log('[changed-file-test-impact-guard] SKIP no changed files detected')\n process.exit(0)\n }\n\n const violations = inspectChangedFiles(changedFiles)\n if (violations.length > 0) {\n fail(violations)\n }\n\n console.log(`[changed-file-test-impact-guard] OK files=${changedFiles.length}`)\n}\n\nconst entryHref = process.argv[1] ? pathToFileURL(process.argv[1]).href : null\nif (entryHref && import.meta.url === entryHref) {\n runCli()\n}\n"
},
{
"path": "scripts/guards/file-line-count-guard.mjs",
"content": "#!/usr/bin/env node\nimport fs from 'fs'\nimport path from 'path'\n\nconst ROOT = process.cwd()\n\nconst RULES = [\n {\n label: 'component',\n dir: 'src',\n include: (relPath) =>\n relPath.includes('/components/')\n && /\\.(ts|tsx)$/.test(relPath),\n limit: 500,\n },\n {\n label: 'hook',\n dir: 'src',\n include: (relPath) =>\n (relPath.includes('/hooks/') || /\\/use[A-Z].+\\.(ts|tsx)$/.test(relPath))\n && /\\.(ts|tsx)$/.test(relPath),\n limit: 400,\n },\n {\n label: 'worker-handler',\n dir: 'src/lib/workers/handlers',\n include: (relPath) => /\\.(ts|tsx)$/.test(relPath),\n limit: 300,\n },\n {\n label: 'mutation',\n dir: 'src/lib/query/mutations',\n include: (relPath) => /\\.(ts|tsx)$/.test(relPath) && !relPath.endsWith('/index.ts'),\n limit: 300,\n },\n]\n\nconst walkFiles = (absDir, relBase = '') => {\n if (!fs.existsSync(absDir)) return []\n const entries = fs.readdirSync(absDir, { withFileTypes: true })\n const out = []\n for (const entry of entries) {\n const abs = path.join(absDir, entry.name)\n const rel = path.join(relBase, entry.name).replace(/\\\\/g, '/')\n if (entry.isDirectory()) {\n out.push(...walkFiles(abs, rel))\n continue\n }\n out.push({ absPath: abs, relPath: rel })\n }\n return out\n}\n\nconst countLines = (absPath) => {\n const raw = fs.readFileSync(absPath, 'utf8')\n if (raw.length === 0) return 0\n return raw.split('\\n').length\n}\n\nconst violations = []\n\nfor (const rule of RULES) {\n const absDir = path.join(ROOT, rule.dir)\n const files = walkFiles(absDir, rule.dir).filter((f) => rule.include(f.relPath))\n for (const file of files) {\n const lineCount = countLines(file.absPath)\n if (lineCount > rule.limit) {\n violations.push({\n label: rule.label,\n relPath: file.relPath,\n lineCount,\n limit: rule.limit,\n })\n }\n }\n}\n\nif (violations.length === 0) {\n process.stdout.write('[file-line-count-guard] PASS\\n')\n process.exit(0)\n}\n\nprocess.stderr.write('[file-line-count-guard] FAIL: file size budget exceeded\\n')\nfor (const violation of violations) {\n process.stderr.write(\n `- [${violation.label}] ${violation.relPath}: ${violation.lineCount} > ${violation.limit}\\n`,\n )\n}\nprocess.exit(1)\n"
},
{
"path": "scripts/guards/image-reference-normalization-guard.mjs",
"content": "#!/usr/bin/env node\n\nimport fs from 'fs'\nimport path from 'path'\nimport process from 'process'\nimport { pathToFileURL } from 'url'\n\nconst root = process.cwd()\nconst handlersDir = path.join(root, 'src', 'lib', 'workers', 'handlers')\n\nexport const NORMALIZATION_HELPER_ALLOWLIST = new Set([\n 'src/lib/workers/handlers/image-task-handler-shared.ts',\n])\n\nconst ACCEPTED_NORMALIZATION_MARKERS = [\n /\\bnormalizeReferenceImagesForGeneration\\s*\\(/,\n /\\bnormalizeToBase64ForGeneration\\s*\\(/,\n /\\bgenerateLabeledImageToCos\\s*\\(/,\n]\n\nfunction fail(title, details = []) {\n process.stderr.write(`\\n[image-reference-normalization-guard] ${title}\\n`)\n for (const detail of details) {\n process.stderr.write(` - ${detail}\\n`)\n }\n process.exit(1)\n}\n\nfunction walk(dir, out = []) {\n if (!fs.existsSync(dir)) return out\n const entries = fs.readdirSync(dir, { withFileTypes: true })\n for (const entry of entries) {\n if (entry.name === '.git' || entry.name === '.next' || entry.name === 'node_modules') continue\n const fullPath = path.join(dir, entry.name)\n if (entry.isDirectory()) {\n walk(fullPath, out)\n continue\n }\n if (entry.name.endsWith('.ts')) out.push(fullPath)\n }\n return out\n}\n\nfunction toRel(fullPath) {\n return path.relative(root, fullPath).split(path.sep).join('/')\n}\n\nfunction usesGenerationReferenceImages(content) {\n return /\\bresolveImageSourceFromGeneration\\s*\\(/.test(content) && /\\breferenceImages\\s*:/.test(content)\n}\n\nfunction hasNormalizationMarker(content) {\n return ACCEPTED_NORMALIZATION_MARKERS.some((pattern) => pattern.test(content))\n}\n\nexport function inspectImageReferenceNormalization(relPath, content) {\n if (NORMALIZATION_HELPER_ALLOWLIST.has(relPath)) return []\n if (!usesGenerationReferenceImages(content)) return []\n if (hasNormalizationMarker(content)) return []\n return [\n `${relPath} uses resolveImageSourceFromGeneration with referenceImages but does not reference normalizeReferenceImagesForGeneration/normalizeToBase64ForGeneration/generateLabeledImageToCos`,\n ]\n}\n\nexport function findImageReferenceNormalizationViolations(scanRoot = root) {\n const scanDir = path.join(scanRoot, 'src', 'lib', 'workers', 'handlers')\n return walk(scanDir)\n .map((fullPath) => {\n const relPath = path.relative(scanRoot, fullPath).split(path.sep).join('/')\n const content = fs.readFileSync(fullPath, 'utf8')\n return inspectImageReferenceNormalization(relPath, content)\n })\n .flat()\n}\n\nexport function main() {\n if (!fs.existsSync(handlersDir)) {\n fail('Missing src/lib/workers/handlers directory')\n }\n\n const handlerFiles = walk(handlersDir)\n const violations = handlerFiles\n .map((fullPath) => {\n const relPath = toRel(fullPath)\n const content = fs.readFileSync(fullPath, 'utf8')\n return inspectImageReferenceNormalization(relPath, content)\n })\n .flat()\n\n if (violations.length > 0) {\n fail('Found image reference normalization violations', violations)\n }\n\n process.stdout.write(\n `[image-reference-normalization-guard] OK handlers=${handlerFiles.length} allowlist=${NORMALIZATION_HELPER_ALLOWLIST.size}\\n`,\n )\n}\n\nif (process.argv[1] && import.meta.url === pathToFileURL(process.argv[1]).href) {\n main()\n}\n"
},
{
"path": "scripts/guards/locale-navigation-guard.mjs",
"content": "#!/usr/bin/env node\n\nimport fs from 'fs'\nimport path from 'path'\nimport process from 'process'\n\nconst root = process.cwd()\nconst sourceExtensions = new Set(['.ts', '.tsx', '.js', '.jsx', '.mjs', '.cjs'])\n\nconst scanDirectories = [\n 'src/app/[locale]',\n]\n\nconst extraFiles = [\n 'src/components/Navbar.tsx',\n 'src/components/LanguageSwitcher.tsx',\n]\n\nfunction toRel(fullPath) {\n return path.relative(root, fullPath).split(path.sep).join('/')\n}\n\nfunction walk(dir, out = []) {\n if (!fs.existsSync(dir)) return out\n const entries = fs.readdirSync(dir, { withFileTypes: true })\n for (const entry of entries) {\n if (entry.name === '.git' || entry.name === '.next' || entry.name === 'node_modules') continue\n const fullPath = path.join(dir, entry.name)\n if (entry.isDirectory()) {\n walk(fullPath, out)\n continue\n }\n if (sourceExtensions.has(path.extname(fullPath))) {\n out.push(fullPath)\n }\n }\n return out\n}\n\nfunction gatherTargetFiles() {\n const files = scanDirectories.flatMap((dir) => walk(path.join(root, dir)))\n for (const relPath of extraFiles) {\n const fullPath = path.join(root, relPath)\n if (fs.existsSync(fullPath)) {\n files.push(fullPath)\n }\n }\n return Array.from(new Set(files))\n}\n\nfunction findViolations(content, relPath) {\n const violations = []\n const lines = content.split('\\n')\n\n const nextLinkImport = /from\\s+['\"]next\\/link['\"]/\n const nextNavigationUseRouterImport = /import\\s*{[\\s\\S]*?\\buseRouter\\b[\\s\\S]*?}\\s*from\\s*['\"]next\\/navigation['\"]/m\n const rootHrefLiteral = /\\bhref\\s*=\\s*[\"']\\//\n const rootHrefTemplate = /\\bhref\\s*=\\s*{`\\//\n const rootRouterCall = /\\brouter\\.(push|replace|prefetch)\\s*\\(\\s*[\"'`]\\//\n\n const nextLinkIndex = content.search(nextLinkImport)\n if (nextLinkIndex >= 0) {\n const lineNo = content.slice(0, nextLinkIndex).split('\\n').length\n violations.push(`${relPath}:${lineNo} do not import next/link in locale navigation surface; use @/i18n/navigation Link`)\n }\n\n const nextNavigationRouterIndex = content.search(nextNavigationUseRouterImport)\n if (nextNavigationRouterIndex >= 0) {\n const lineNo = content.slice(0, nextNavigationRouterIndex).split('\\n').length\n violations.push(`${relPath}:${lineNo} do not import useRouter from next/navigation in locale navigation surface; use @/i18n/navigation useRouter`)\n }\n\n for (let index = 0; index < lines.length; index += 1) {\n const line = lines[index]\n const lineNo = index + 1\n if (rootHrefLiteral.test(line) || rootHrefTemplate.test(line)) {\n violations.push(`${relPath}:${lineNo} do not use root-literal href; use Link href={{ pathname: '...' }} via @/i18n/navigation`)\n }\n if (rootRouterCall.test(line)) {\n violations.push(`${relPath}:${lineNo} do not use root-literal router navigation; use router.push/replace({ pathname: '...' }) via @/i18n/navigation`)\n }\n }\n\n return violations\n}\n\nconst violations = []\nfor (const filePath of gatherTargetFiles()) {\n const content = fs.readFileSync(filePath, 'utf8')\n violations.push(...findViolations(content, toRel(filePath)))\n}\n\nif (violations.length > 0) {\n console.error('\\n[locale-navigation-guard] violations found:')\n for (const violation of violations) {\n console.error(` - ${violation}`)\n }\n process.exit(1)\n}\n\nconsole.log('[locale-navigation-guard] OK')\n"
},
{
"path": "scripts/guards/no-api-direct-llm-call.mjs",
"content": "#!/usr/bin/env node\n\nimport fs from 'fs'\nimport path from 'path'\nimport process from 'process'\n\nconst root = process.cwd()\nconst scanRoots = ['src/app/api', 'src/pages/api']\nconst allowedPrefixes = []\nconst sourceExtensions = new Set(['.ts', '.tsx', '.js', '.jsx', '.mjs', '.cjs'])\n\nfunction fail(title, details = []) {\n console.error(`\\n[no-api-direct-llm-call] ${title}`)\n for (const line of details) {\n console.error(` - ${line}`)\n }\n process.exit(1)\n}\n\nfunction toRel(fullPath) {\n return path.relative(root, fullPath).split(path.sep).join('/')\n}\n\nfunction walk(dir, out = []) {\n if (!fs.existsSync(dir)) return out\n const entries = fs.readdirSync(dir, { withFileTypes: true })\n for (const entry of entries) {\n if (entry.name === '.git' || entry.name === '.next' || entry.name === 'node_modules') continue\n const fullPath = path.join(dir, entry.name)\n if (entry.isDirectory()) {\n walk(fullPath, out)\n continue\n }\n const ext = path.extname(entry.name)\n if (sourceExtensions.has(ext)) {\n out.push(fullPath)\n }\n }\n return out\n}\n\nfunction isAllowedFile(relPath) {\n return allowedPrefixes.some((prefix) => relPath.startsWith(prefix))\n}\n\nfunction collectViolations(fullPath) {\n const relPath = toRel(fullPath)\n if (isAllowedFile(relPath)) return []\n\n const content = fs.readFileSync(fullPath, 'utf8')\n const lines = content.split('\\n')\n const violations = []\n\n for (let i = 0; i < lines.length; i += 1) {\n const line = lines[i]\n if (/from\\s+['\"]@\\/lib\\/llm-client['\"]/.test(line)) {\n violations.push(`${relPath}:${i + 1} forbidden import from '@/lib/llm-client'`)\n }\n if (/\\bchatCompletion[A-Za-z0-9_]*\\s*\\(/.test(line)) {\n violations.push(`${relPath}:${i + 1} forbidden direct chatCompletion* call`)\n }\n if (/\\bisInternalTaskExecution\\b/.test(line)) {\n violations.push(`${relPath}:${i + 1} forbidden dual-track fallback marker isInternalTaskExecution`)\n }\n }\n\n return violations\n}\n\nconst allFiles = scanRoots.flatMap((scanRoot) => walk(path.join(root, scanRoot)))\nconst violations = allFiles.flatMap((fullPath) => collectViolations(fullPath))\n\nif (violations.length > 0) {\n fail('Found forbidden direct LLM execution in production API routes', violations)\n}\n\nconsole.log('[no-api-direct-llm-call] OK')\n"
},
{
"path": "scripts/guards/no-duplicate-endpoint-entry.mjs",
"content": "#!/usr/bin/env node\nimport fs from 'fs'\nimport path from 'path'\n\nconst ROOT = process.cwd()\nconst API_ROOT = path.join(ROOT, 'src', 'app', 'api')\n\nconst KNOWN_DUPLICATE_GROUPS = [\n {\n key: 'user-llm-test-connection',\n candidates: [\n 'src/app/api/user/api-config/test-connection/route.ts',\n 'src/app/api/user/test-llm-provider/route.ts',\n ],\n },\n]\n\nconst exists = (relPath) => fs.existsSync(path.join(ROOT, relPath))\n\nconst failures = []\nfor (const group of KNOWN_DUPLICATE_GROUPS) {\n const present = group.candidates.filter(exists)\n if (present.length > 1) {\n failures.push({ key: group.key, present })\n }\n}\n\nif (!fs.existsSync(API_ROOT)) {\n process.stdout.write('[no-duplicate-endpoint-entry] PASS (api dir missing)\\n')\n process.exit(0)\n}\n\nif (failures.length === 0) {\n process.stdout.write('[no-duplicate-endpoint-entry] PASS\\n')\n process.exit(0)\n}\n\nprocess.stderr.write('[no-duplicate-endpoint-entry] FAIL: duplicated endpoint entry detected\\n')\nfor (const failure of failures) {\n process.stderr.write(`- ${failure.key}\\n`)\n for (const relPath of failure.present) {\n process.stderr.write(` - ${relPath}\\n`)\n }\n}\nprocess.exit(1)\n"
},
{
"path": "scripts/guards/no-hardcoded-model-capabilities.mjs",
"content": "#!/usr/bin/env node\n\nimport fs from 'fs'\nimport path from 'path'\nimport process from 'process'\n\nconst root = process.cwd()\nconst sourceExtensions = new Set(['.ts', '.tsx', '.js', '.jsx', '.mjs', '.cjs'])\nconst scanRoots = ['src']\nconst allowConstantDefinitionsIn = new Set([\n 'src/lib/constants.ts',\n])\nconst forbiddenCapabilityConstants = [\n 'VIDEO_MODELS',\n 'FIRST_LAST_FRAME_MODELS',\n 'AUDIO_SUPPORTED_MODELS',\n 'BANANA_MODELS',\n 'BANANA_RESOLUTION_OPTIONS',\n]\n\nfunction fail(title, details = []) {\n console.error(`\\n[no-hardcoded-model-capabilities] ${title}`)\n for (const line of details) {\n console.error(` - ${line}`)\n }\n process.exit(1)\n}\n\nfunction toRel(fullPath) {\n return path.relative(root, fullPath).split(path.sep).join('/')\n}\n\nfunction walk(dir, out = []) {\n if (!fs.existsSync(dir)) return out\n const entries = fs.readdirSync(dir, { withFileTypes: true })\n for (const entry of entries) {\n if (entry.name === '.git' || entry.name === '.next' || entry.name === 'node_modules') continue\n const fullPath = path.join(dir, entry.name)\n if (entry.isDirectory()) {\n walk(fullPath, out)\n continue\n }\n if (sourceExtensions.has(path.extname(entry.name))) {\n out.push(fullPath)\n }\n }\n return out\n}\n\nconst files = scanRoots.flatMap((scanRoot) => walk(path.join(root, scanRoot)))\nconst violations = []\n\nfor (const fullPath of files) {\n const relPath = toRel(fullPath)\n if (allowConstantDefinitionsIn.has(relPath)) continue\n\n const lines = fs.readFileSync(fullPath, 'utf8').split('\\n')\n for (let index = 0; index < lines.length; index += 1) {\n const line = lines[index]\n for (const token of forbiddenCapabilityConstants) {\n const tokenPattern = new RegExp(`\\\\b${token}\\\\b`)\n if (tokenPattern.test(line)) {\n violations.push(`${relPath}:${index + 1} forbidden hardcoded model capability token ${token}`)\n }\n }\n }\n}\n\nif (violations.length > 0) {\n fail('Found hardcoded model capability usage', violations)\n}\n\nconsole.log('[no-hardcoded-model-capabilities] OK')\n"
},
{
"path": "scripts/guards/no-internal-task-sync-fallback.mjs",
"content": "#!/usr/bin/env node\n\nimport fs from 'fs'\nimport path from 'path'\nimport process from 'process'\n\nconst root = process.cwd()\nconst scanRoots = ['src/app/api', 'src/pages/api']\nconst allowedPrefixes = ['src/app/api/ui-review/', 'src/pages/api/ui-review/']\nconst sourceExtensions = new Set(['.ts', '.tsx', '.js', '.jsx', '.mjs', '.cjs'])\n\nfunction fail(title, details = []) {\n console.error(`\\n[no-internal-task-sync-fallback] ${title}`)\n for (const line of details) {\n console.error(` - ${line}`)\n }\n process.exit(1)\n}\n\nfunction toRel(fullPath) {\n return path.relative(root, fullPath).split(path.sep).join('/')\n}\n\nfunction walk(dir, out = []) {\n if (!fs.existsSync(dir)) return out\n const entries = fs.readdirSync(dir, { withFileTypes: true })\n for (const entry of entries) {\n if (entry.name === '.git' || entry.name === '.next' || entry.name === 'node_modules') continue\n const fullPath = path.join(dir, entry.name)\n if (entry.isDirectory()) {\n walk(fullPath, out)\n continue\n }\n if (sourceExtensions.has(path.extname(entry.name))) {\n out.push(fullPath)\n }\n }\n return out\n}\n\nfunction isAllowedFile(relPath) {\n return allowedPrefixes.some((prefix) => relPath.startsWith(prefix))\n}\n\nfunction collectViolations(fullPath) {\n const relPath = toRel(fullPath)\n if (isAllowedFile(relPath)) return []\n\n const content = fs.readFileSync(fullPath, 'utf8')\n const lines = content.split('\\n')\n const violations = []\n\n for (let i = 0; i < lines.length; i += 1) {\n const line = lines[i]\n if (/\\bisInternalTaskExecution\\b/.test(line)) {\n violations.push(`${relPath}:${i + 1} forbidden dual-track fallback marker isInternalTaskExecution`)\n }\n if (/\\bshouldRunSyncTask\\s*\\(/.test(line)) {\n violations.push(`${relPath}:${i + 1} forbidden sync-mode branch helper shouldRunSyncTask`)\n }\n }\n\n if (/\\bmaybeSubmitLLMTask\\s*\\(/.test(content) && !/sync mode is disabled for this route/.test(content)) {\n violations.push(`${relPath} missing explicit sync-disabled guard after maybeSubmitLLMTask`)\n }\n\n return violations\n}\n\nconst allFiles = scanRoots.flatMap((scanRoot) => walk(path.join(root, scanRoot)))\nconst violations = allFiles.flatMap((fullPath) => collectViolations(fullPath))\n\nif (violations.length > 0) {\n fail('Found potential sync fallback or dual-track task branch in production API routes', violations)\n}\n\nconsole.log('[no-internal-task-sync-fallback] OK')\n"
},
{
"path": "scripts/guards/no-media-provider-bypass.mjs",
"content": "#!/usr/bin/env node\n\nimport fs from 'fs'\nimport path from 'path'\nimport process from 'process'\n\nconst root = process.cwd()\nconst sourceExtensions = new Set(['.ts', '.tsx', '.js', '.jsx', '.mjs', '.cjs'])\nconst allowFactoryImportIn = new Set([\n 'src/lib/generator-api.ts',\n 'src/lib/generators/factory.ts',\n])\n\nfunction fail(title, details = []) {\n console.error(`\\n[no-media-provider-bypass] ${title}`)\n for (const line of details) {\n console.error(` - ${line}`)\n }\n process.exit(1)\n}\n\nfunction toRel(fullPath) {\n return path.relative(root, fullPath).split(path.sep).join('/')\n}\n\nfunction walk(dir, out = []) {\n if (!fs.existsSync(dir)) return out\n const entries = fs.readdirSync(dir, { withFileTypes: true })\n for (const entry of entries) {\n if (entry.name === '.git' || entry.name === '.next' || entry.name === 'node_modules') continue\n const fullPath = path.join(dir, entry.name)\n if (entry.isDirectory()) {\n walk(fullPath, out)\n continue\n }\n if (sourceExtensions.has(path.extname(entry.name))) {\n out.push(fullPath)\n }\n }\n return out\n}\n\nconst generatorApiPath = path.join(root, 'src/lib/generator-api.ts')\nif (!fs.existsSync(generatorApiPath)) {\n fail('Missing src/lib/generator-api.ts')\n}\n\nconst generatorApiContent = fs.readFileSync(generatorApiPath, 'utf8')\nconst resolveModelSelectionHits = (generatorApiContent.match(/resolveModelSelection\\s*\\(/g) || []).length\nif (resolveModelSelectionHits < 2) {\n fail('generator-api must route both image and video generation through resolveModelSelection', [\n 'expected >= 2 resolveModelSelection(...) calls in src/lib/generator-api.ts',\n ])\n}\n\nconst allFiles = walk(path.join(root, 'src'))\nconst violations = []\n\nfor (const fullPath of allFiles) {\n const relPath = toRel(fullPath)\n const content = fs.readFileSync(fullPath, 'utf8')\n const lines = content.split('\\n')\n\n for (let i = 0; i < lines.length; i += 1) {\n const line = lines[i]\n\n if (\n relPath !== 'src/lib/generators/factory.ts' &&\n (/\\bcreateImageGeneratorByModel\\s*\\(/.test(line) || /\\bcreateVideoGeneratorByModel\\s*\\(/.test(line))\n ) {\n violations.push(`${relPath}:${i + 1} forbidden provider-bypass factory call create*GeneratorByModel(...)`)\n }\n\n if ((/\\bgetImageApiKey\\s*\\(/.test(line) || /\\bgetVideoApiKey\\s*\\(/.test(line)) && relPath !== 'src/lib/api-config.ts') {\n violations.push(`${relPath}:${i + 1} forbidden direct getImageApiKey/getVideoApiKey usage outside api-config`)\n }\n\n if (/from\\s+['\"]@\\/lib\\/generators\\/factory['\"]/.test(line) && !allowFactoryImportIn.has(relPath)) {\n violations.push(`${relPath}:${i + 1} forbidden direct import from '@/lib/generators/factory' (must go through generator-api)`)\n }\n }\n}\n\nif (violations.length > 0) {\n fail('Found media provider routing bypass', violations)\n}\n\nconsole.log('[no-media-provider-bypass] OK')\n"
},
{
"path": "scripts/guards/no-model-key-downgrade.mjs",
"content": "#!/usr/bin/env node\n\nimport fs from 'fs'\nimport path from 'path'\nimport process from 'process'\n\nconst root = process.cwd()\nconst sourceExtensions = new Set(['.ts', '.tsx', '.js', '.jsx', '.mjs', '.cjs'])\nconst scanRoots = ['src/app', 'src/lib']\nconst modelFields = [\n 'analysisModel',\n 'characterModel',\n 'locationModel',\n 'storyboardModel',\n 'editModel',\n 'videoModel',\n]\n\nfunction fail(title, details = []) {\n console.error(`\\n[no-model-key-downgrade] ${title}`)\n for (const line of details) {\n console.error(` - ${line}`)\n }\n process.exit(1)\n}\n\nfunction toRel(fullPath) {\n return path.relative(root, fullPath).split(path.sep).join('/')\n}\n\nfunction walk(dir, out = []) {\n if (!fs.existsSync(dir)) return out\n const entries = fs.readdirSync(dir, { withFileTypes: true })\n for (const entry of entries) {\n if (entry.name === '.git' || entry.name === '.next' || entry.name === 'node_modules') continue\n const fullPath = path.join(dir, entry.name)\n if (entry.isDirectory()) {\n walk(fullPath, out)\n continue\n }\n if (sourceExtensions.has(path.extname(entry.name))) {\n out.push(fullPath)\n }\n }\n return out\n}\n\nfunction collectViolations(filePath) {\n const relPath = toRel(filePath)\n const lines = fs.readFileSync(filePath, 'utf8').split('\\n')\n const violations = []\n\n const modelFieldPattern = new RegExp(`\\\\b(${modelFields.join('|')})\\\\s*:\\\\s*[^,\\\\n]*\\\\bmodelId\\\\b`)\n const optionModelIdPattern = /value=\\{model\\.modelId\\}/\n\n for (let index = 0; index < lines.length; index += 1) {\n const line = lines[index]\n if (modelFieldPattern.test(line)) {\n violations.push(`${relPath}:${index + 1} default model field must persist model_key, not modelId`)\n }\n if (optionModelIdPattern.test(line)) {\n violations.push(`${relPath}:${index + 1} UI option value must use modelKey, not model.modelId`)\n }\n }\n\n return violations\n}\n\nfunction assertFileContains(relativePath, requiredSnippets) {\n const fullPath = path.join(root, relativePath)\n if (!fs.existsSync(fullPath)) {\n fail('Missing required contract file', [relativePath])\n }\n const content = fs.readFileSync(fullPath, 'utf8')\n const missing = requiredSnippets.filter((snippet) => !content.includes(snippet))\n if (missing.length > 0) {\n fail('Model key contract anchor missing', missing.map((snippet) => `${relativePath} missing: ${snippet}`))\n }\n}\n\nconst files = scanRoots.flatMap((scanRoot) => walk(path.join(root, scanRoot)))\nconst violations = files.flatMap((filePath) => collectViolations(filePath))\n\nassertFileContains('src/lib/model-config-contract.ts', ['parseModelKeyStrict', 'markerIndex === -1) return null'])\nassertFileContains('src/lib/config-service.ts', ['parseModelKeyStrict'])\nassertFileContains('src/app/api/user/api-config/route.ts', ['validateDefaultModelKey', 'must be provider::modelId'])\nassertFileContains('src/app/api/novel-promotion/[projectId]/route.ts', ['must be provider::modelId'])\n\nif (violations.length > 0) {\n fail('Found model key downgrade pattern', violations)\n}\n\nconsole.log('[no-model-key-downgrade] OK')\n"
},
{
"path": "scripts/guards/no-multiple-sources-of-truth.mjs",
"content": "#!/usr/bin/env node\n\nimport fs from 'fs'\nimport path from 'path'\nimport process from 'process'\n\nconst root = process.cwd()\nconst sourceExtensions = new Set(['.ts', '.tsx', '.js', '.jsx', '.mjs', '.cjs'])\n\nconst lineScanRoots = [\n 'src/app/[locale]/workspace/[projectId]/modes/novel-promotion',\n 'src/lib/query/hooks',\n]\n\nconst fileScanRoots = [\n 'src/app/api/novel-promotion',\n 'src/lib/workers/handlers',\n]\n\nconst lineRules = [\n {\n name: 'shadow state localStoryboards',\n test: (line) => /const\\s*\\[\\s*localStoryboards\\s*,\\s*setLocalStoryboards\\s*\\]\\s*=\\s*useState/.test(line),\n },\n {\n name: 'shadow state localVoiceLines',\n test: (line) => /const\\s*\\[\\s*localVoiceLines\\s*,\\s*setLocalVoiceLines\\s*\\]\\s*=\\s*useState/.test(line),\n },\n {\n name: 'hardcoded queryKey array',\n test: (line) => /queryKey\\s*:\\s*\\[/.test(line),\n },\n]\n\nfunction fail(title, details = []) {\n console.error(`\\n[no-multiple-sources-of-truth] ${title}`)\n for (const detail of details) {\n console.error(` - ${detail}`)\n }\n process.exit(1)\n}\n\nfunction toRel(fullPath) {\n return path.relative(root, fullPath).split(path.sep).join('/')\n}\n\nfunction walk(dir, out = []) {\n if (!fs.existsSync(dir)) return out\n const entries = fs.readdirSync(dir, { withFileTypes: true })\n for (const entry of entries) {\n if (entry.name === '.git' || entry.name === '.next' || entry.name === 'node_modules') continue\n const fullPath = path.join(dir, entry.name)\n if (entry.isDirectory()) {\n walk(fullPath, out)\n continue\n }\n if (sourceExtensions.has(path.extname(entry.name))) out.push(fullPath)\n }\n return out\n}\n\nfunction collectLineViolations(fullPath) {\n const relPath = toRel(fullPath)\n const content = fs.readFileSync(fullPath, 'utf8')\n const lines = content.split('\\n')\n const violations = []\n\n for (let i = 0; i < lines.length; i += 1) {\n const line = lines[i]\n for (const rule of lineRules) {\n if (rule.test(line)) {\n violations.push(`${relPath}:${i + 1} forbidden: ${rule.name}`)\n }\n }\n }\n\n return violations\n}\n\nfunction collectFileViolations(fullPath) {\n const relPath = toRel(fullPath)\n const content = fs.readFileSync(fullPath, 'utf8')\n const violations = []\n\n const updateCallRegex = /novelPromotionProject\\.update\\(\\{[\\s\\S]*?\\n\\s*\\}\\)/g\n for (const match of content.matchAll(updateCallRegex)) {\n const block = match[0]\n const hasStageWrite = /\\bdata\\s*:\\s*\\{[\\s\\S]*?\\bstage\\s*:/.test(block)\n if (!hasStageWrite) continue\n const before = content.slice(0, match.index ?? 0)\n const lineNumber = before.split('\\n').length\n violations.push(`${relPath}:${lineNumber} forbidden: DB stage write in novelPromotionProject.update`)\n }\n\n return violations\n}\n\nconst lineFiles = lineScanRoots.flatMap((scanRoot) => walk(path.join(root, scanRoot)))\nconst fileFiles = fileScanRoots.flatMap((scanRoot) => walk(path.join(root, scanRoot)))\n\nconst lineViolations = lineFiles.flatMap((fullPath) => collectLineViolations(fullPath))\nconst fileViolations = fileFiles.flatMap((fullPath) => collectFileViolations(fullPath))\nconst allViolations = [...lineViolations, ...fileViolations]\n\nif (allViolations.length > 0) {\n fail('Found multiple-sources-of-truth regressions', allViolations)\n}\n\nconsole.log('[no-multiple-sources-of-truth] OK')\n"
},
{
"path": "scripts/guards/no-provider-guessing.mjs",
"content": "#!/usr/bin/env node\n\nimport fs from 'fs'\nimport path from 'path'\nimport process from 'process'\n\nconst root = process.cwd()\nconst sourceExtensions = new Set(['.ts', '.tsx', '.js', '.jsx', '.mjs', '.cjs'])\nconst scanRoots = ['src/lib', 'src/app/api']\nconst allowModelRegistryUsage = new Set()\n\nfunction fail(title, details = []) {\n console.error(`\\n[no-provider-guessing] ${title}`)\n for (const line of details) {\n console.error(` - ${line}`)\n }\n process.exit(1)\n}\n\nfunction toRel(fullPath) {\n return path.relative(root, fullPath).split(path.sep).join('/')\n}\n\nfunction walk(dir, out = []) {\n if (!fs.existsSync(dir)) return out\n const entries = fs.readdirSync(dir, { withFileTypes: true })\n for (const entry of entries) {\n if (entry.name === '.git' || entry.name === '.next' || entry.name === 'node_modules') continue\n const fullPath = path.join(dir, entry.name)\n if (entry.isDirectory()) {\n walk(fullPath, out)\n continue\n }\n if (sourceExtensions.has(path.extname(entry.name))) {\n out.push(fullPath)\n }\n }\n return out\n}\n\nconst apiConfigPath = path.join(root, 'src/lib/api-config.ts')\nif (!fs.existsSync(apiConfigPath)) {\n fail('Missing src/lib/api-config.ts')\n}\nconst legacyRegistryPath = path.join(root, 'src/lib/model-registry.ts')\nif (fs.existsSync(legacyRegistryPath)) {\n fail('Legacy runtime registry must be removed', ['src/lib/model-registry.ts'])\n}\nconst apiConfigText = fs.readFileSync(apiConfigPath, 'utf8')\n\nconst forbiddenApiConfigTokens = [\n 'includeAnyType',\n 'crossTypeCandidates',\n 'matches multiple providers across media types',\n]\nconst apiViolations = forbiddenApiConfigTokens\n .filter((token) => apiConfigText.includes(token))\n .map((token) => `src/lib/api-config.ts contains forbidden provider-guessing token: ${token}`)\n\n// 验证 api-config.ts 使用严格 provider.id 精确匹配（不按 type 过滤，不做 providerKey 模糊匹配）\nif (!apiConfigText.includes('pickProviderStrict(')) {\n apiViolations.push('src/lib/api-config.ts missing strict provider resolution function (pickProviderStrict)')\n}\n\nconst files = scanRoots.flatMap((scanRoot) => walk(path.join(root, scanRoot)))\nconst violations = [...apiViolations]\n\nfor (const fullPath of files) {\n const relPath = toRel(fullPath)\n const content = fs.readFileSync(fullPath, 'utf8')\n const lines = content.split('\\n')\n\n for (let index = 0; index < lines.length; index += 1) {\n const line = lines[index]\n if (\n /from\\s+['\"]@\\/lib\\/model-registry['\"]/.test(line)\n && !allowModelRegistryUsage.has(relPath)\n ) {\n violations.push(`${relPath}:${index + 1} forbidden model-registry import outside allowed boundary`)\n }\n\n if (\n (/\\bgetModelRegistryEntry\\s*\\(/.test(line) || /\\blistRegisteredModels\\s*\\(/.test(line))\n && !allowModelRegistryUsage.has(relPath)\n ) {\n violations.push(`${relPath}:${index + 1} forbidden model-registry runtime mapping usage`)\n }\n }\n}\n\nif (violations.length > 0) {\n fail('Found provider guessing / registry mapping violation', violations)\n}\n\nconsole.log('[no-provider-guessing] OK')\n"
},
{
"path": "scripts/guards/no-server-mirror-state.mjs",
"content": "#!/usr/bin/env node\n\nimport fs from 'fs'\nimport path from 'path'\nimport process from 'process'\n\nconst root = process.cwd()\nconst scanRoots = [\n 'src/app/[locale]/workspace/[projectId]/modes/novel-promotion',\n]\nconst sourceExtensions = new Set(['.ts', '.tsx', '.js', '.jsx', '.mjs', '.cjs'])\n\nconst forbiddenRules = [\n {\n name: 'localProject/localEpisode mirror state',\n test: (line) => /\\blocalProject\\b|\\blocalEpisode\\b/.test(line),\n },\n {\n name: 'server mirror useState(projectData.*)',\n test: (line) => /useState\\s*\\(\\s*projectData\\./.test(line),\n },\n {\n name: 'server mirror useState(episode?.*)',\n test: (line) => /useState\\s*\\(\\s*episode\\?\\./.test(line),\n },\n]\n\nfunction fail(title, details = []) {\n console.error(`\\n[no-server-mirror-state] ${title}`)\n for (const line of details) {\n console.error(` - ${line}`)\n }\n process.exit(1)\n}\n\nfunction toRel(fullPath) {\n return path.relative(root, fullPath).split(path.sep).join('/')\n}\n\nfunction walk(dir, out = []) {\n if (!fs.existsSync(dir)) return out\n const entries = fs.readdirSync(dir, { withFileTypes: true })\n for (const entry of entries) {\n if (entry.name === '.git' || entry.name === '.next' || entry.name === 'node_modules') continue\n const fullPath = path.join(dir, entry.name)\n if (entry.isDirectory()) {\n walk(fullPath, out)\n continue\n }\n const ext = path.extname(entry.name)\n if (sourceExtensions.has(ext)) out.push(fullPath)\n }\n return out\n}\n\nfunction collectViolations(fullPath) {\n const relPath = toRel(fullPath)\n const content = fs.readFileSync(fullPath, 'utf8')\n const lines = content.split('\\n')\n const violations = []\n\n for (let i = 0; i < lines.length; i += 1) {\n const line = lines[i]\n for (const rule of forbiddenRules) {\n if (rule.test(line)) {\n violations.push(`${relPath}:${i + 1} forbidden: ${rule.name}`)\n }\n }\n }\n\n return violations\n}\n\nconst allFiles = scanRoots.flatMap((scanRoot) => walk(path.join(root, scanRoot)))\nconst violations = allFiles.flatMap((fullPath) => collectViolations(fullPath))\n\nif (violations.length > 0) {\n fail('Found forbidden server mirror state patterns', violations)\n}\n\nconsole.log('[no-server-mirror-state] OK')\n"
},
{
"path": "scripts/guards/prompt-ab-regression.mjs",
"content": "#!/usr/bin/env node\n\nimport fs from 'fs'\nimport path from 'path'\nimport process from 'process'\n\nconst root = process.cwd()\nconst catalogPath = path.join(root, 'src', 'lib', 'prompt-i18n', 'catalog.ts')\nconst singlePlaceholderPattern = /\\{([A-Za-z0-9_]+)\\}/g\nconst doublePlaceholderPattern = /\\{\\{([A-Za-z0-9_]+)\\}\\}/g\nconst unresolvedPlaceholderPattern = /\\{\\{?[A-Za-z0-9_]+\\}?\\}/g\n\nfunction fail(title, details = []) {\n console.error(`\\n[prompt-ab-regression] ${title}`)\n for (const line of details) {\n console.error(` - ${line}`)\n }\n process.exit(1)\n}\n\nfunction parseCatalog(text) {\n const entries = []\n const entryPattern = /pathStem:\\s*'([^']+)'\\s*,[\\s\\S]*?variableKeys:\\s*\\[([\\s\\S]*?)\\]\\s*,/g\n for (const match of text.matchAll(entryPattern)) {\n const pathStem = match[1]\n const rawKeys = match[2] || ''\n const keys = Array.from(rawKeys.matchAll(/'([^']+)'/g)).map((item) => item[1])\n entries.push({ pathStem, variableKeys: keys })\n }\n return entries\n}\n\nfunction extractPlaceholders(template) {\n const keys = new Set()\n for (const match of template.matchAll(singlePlaceholderPattern)) {\n if (match[1]) keys.add(match[1])\n }\n for (const match of template.matchAll(doublePlaceholderPattern)) {\n if (match[1]) keys.add(match[1])\n }\n return Array.from(keys)\n}\n\nfunction replaceAll(template, variables) {\n let rendered = template\n for (const [key, value] of Object.entries(variables)) {\n const pattern = new RegExp(`\\\\{\\\\{${key}\\\\}\\\\}|\\\\{${key}\\\\}`, 'g')\n rendered = rendered.replace(pattern, value)\n }\n return rendered\n}\n\nfunction setDiff(left, right) {\n const rightSet = new Set(right)\n return left.filter((item) => !rightSet.has(item))\n}\n\nif (!fs.existsSync(catalogPath)) {\n fail('catalog.ts not found', ['src/lib/prompt-i18n/catalog.ts'])\n}\n\nconst catalogText = fs.readFileSync(catalogPath, 'utf8')\nconst entries = parseCatalog(catalogText)\nif (entries.length === 0) {\n fail('failed to parse prompt catalog entries')\n}\n\nconst violations = []\n\nfor (const entry of entries) {\n const zhPath = path.join(root, 'lib', 'prompts', `${entry.pathStem}.zh.txt`)\n const enPath = path.join(root, 'lib', 'prompts', `${entry.pathStem}.en.txt`)\n if (!fs.existsSync(zhPath)) {\n violations.push(`missing zh template: lib/prompts/${entry.pathStem}.zh.txt`)\n continue\n }\n if (!fs.existsSync(enPath)) {\n violations.push(`missing en template: lib/prompts/${entry.pathStem}.en.txt`)\n continue\n }\n\n const zhTemplate = fs.readFileSync(zhPath, 'utf8')\n const enTemplate = fs.readFileSync(enPath, 'utf8')\n const declared = entry.variableKeys\n const zhPlaceholders = extractPlaceholders(zhTemplate)\n const enPlaceholders = extractPlaceholders(enTemplate)\n\n const missingInZh = setDiff(declared, zhPlaceholders)\n const missingInEn = setDiff(declared, enPlaceholders)\n const extraInZh = setDiff(zhPlaceholders, declared)\n const extraInEn = setDiff(enPlaceholders, declared)\n const zhOnly = setDiff(zhPlaceholders, enPlaceholders)\n const enOnly = setDiff(enPlaceholders, zhPlaceholders)\n\n for (const key of missingInZh) {\n violations.push(`missing {${key}} in zh template: lib/prompts/${entry.pathStem}.zh.txt`)\n }\n for (const key of missingInEn) {\n violations.push(`missing {${key}} in en template: lib/prompts/${entry.pathStem}.en.txt`)\n }\n for (const key of extraInZh) {\n violations.push(`unexpected {${key}} in zh template: lib/prompts/${entry.pathStem}.zh.txt`)\n }\n for (const key of extraInEn) {\n violations.push(`unexpected {${key}} in en template: lib/prompts/${entry.pathStem}.en.txt`)\n }\n for (const key of zhOnly) {\n violations.push(`placeholder {${key}} exists only in zh template: ${entry.pathStem}`)\n }\n for (const key of enOnly) {\n violations.push(`placeholder {${key}} exists only in en template: ${entry.pathStem}`)\n }\n\n const variables = Object.fromEntries(\n declared.map((key) => [key, `__AB_SAMPLE_${key.toUpperCase()}__`]),\n )\n const renderedZh = replaceAll(zhTemplate, variables)\n const renderedEn = replaceAll(enTemplate, variables)\n\n const unresolvedZh = renderedZh.match(unresolvedPlaceholderPattern) || []\n const unresolvedEn = renderedEn.match(unresolvedPlaceholderPattern) || []\n if (unresolvedZh.length > 0) {\n violations.push(`unresolved placeholders in zh template: ${entry.pathStem} -> ${unresolvedZh.join(', ')}`)\n }\n if (unresolvedEn.length > 0) {\n violations.push(`unresolved placeholders in en template: ${entry.pathStem} -> ${unresolvedEn.join(', ')}`)\n }\n\n for (const [key, sample] of Object.entries(variables)) {\n if (!renderedZh.includes(sample)) {\n violations.push(`zh template variable not used after render: ${entry.pathStem}.{${key}}`)\n }\n if (!renderedEn.includes(sample)) {\n violations.push(`en template variable not used after render: ${entry.pathStem}.{${key}}`)\n }\n }\n}\n\nif (violations.length > 0) {\n fail('A/B regression check failed', violations)\n}\n\nconsole.log(`[prompt-ab-regression] OK (${entries.length} templates checked)`)\n"
},
{
"path": "scripts/guards/prompt-i18n-guard.mjs",
"content": "#!/usr/bin/env node\n\nimport fs from 'fs'\nimport path from 'path'\nimport process from 'process'\n\nconst root = process.cwd()\nconst sourceExtensions = new Set(['.ts', '.tsx', '.js', '.jsx', '.mjs', '.cjs'])\nconst scanRoots = ['src', 'scripts']\nconst allowedPromptTemplateReaders = new Set([\n 'src/lib/prompt-i18n/template-store.ts',\n 'scripts/guards/prompt-i18n-guard.mjs',\n 'scripts/guards/prompt-semantic-regression.mjs',\n 'scripts/guards/prompt-ab-regression.mjs',\n 'scripts/guards/prompt-json-canary-guard.mjs',\n])\nconst languageDirectiveAllowList = new Set([\n 'scripts/guards/prompt-i18n-guard.mjs',\n])\nconst languageDirectivePattern = /请用中文|中文输出|use Chinese|output in Chinese/i\n\nfunction fail(title, details = []) {\n console.error(`\\n[prompt-i18n-guard] ${title}`)\n for (const line of details) {\n console.error(` - ${line}`)\n }\n process.exit(1)\n}\n\nfunction toRel(fullPath) {\n return path.relative(root, fullPath).split(path.sep).join('/')\n}\n\nfunction walk(dir, out = []) {\n if (!fs.existsSync(dir)) return out\n const entries = fs.readdirSync(dir, { withFileTypes: true })\n for (const entry of entries) {\n if (entry.name === '.git' || entry.name === '.next' || entry.name === 'node_modules') continue\n const fullPath = path.join(dir, entry.name)\n if (entry.isDirectory()) {\n walk(fullPath, out)\n continue\n }\n out.push(fullPath)\n }\n return out\n}\n\nfunction listSourceFiles() {\n return scanRoots\n .flatMap((scanRoot) => walk(path.join(root, scanRoot)))\n .filter((fullPath) => sourceExtensions.has(path.extname(fullPath)))\n}\n\nfunction collectDirectPromptReadViolations() {\n const violations = []\n const files = listSourceFiles()\n for (const filePath of files) {\n const relPath = toRel(filePath)\n if (allowedPromptTemplateReaders.has(relPath)) continue\n const content = fs.readFileSync(filePath, 'utf8')\n const hasReadFileSync = /\\breadFileSync\\s*\\(/.test(content)\n if (!hasReadFileSync) continue\n const hasPromptPathToken =\n content.includes('lib/prompts')\n || (\n /['\"]lib['\"]/.test(content)\n && /['\"]prompts['\"]/.test(content)\n )\n if (hasPromptPathToken) {\n violations.push(`${relPath} direct prompt file read is forbidden; use buildPrompt/getPromptTemplate`)\n }\n }\n return violations\n}\n\nfunction collectLanguageDirectiveViolations() {\n const violations = []\n\n for (const filePath of listSourceFiles()) {\n const relPath = toRel(filePath)\n if (languageDirectiveAllowList.has(relPath)) continue\n const lines = fs.readFileSync(filePath, 'utf8').split('\\n')\n for (let index = 0; index < lines.length; index += 1) {\n const line = lines[index]\n if (languageDirectivePattern.test(line)) {\n violations.push(`${relPath}:${index + 1} hardcoded language directive is forbidden`)\n }\n }\n }\n\n const promptFiles = walk(path.join(root, 'lib', 'prompts'))\n .filter((fullPath) => fullPath.endsWith('.en.txt'))\n for (const filePath of promptFiles) {\n const relPath = toRel(filePath)\n const lines = fs.readFileSync(filePath, 'utf8').split('\\n')\n for (let index = 0; index < lines.length; index += 1) {\n const line = lines[index]\n if (languageDirectivePattern.test(line)) {\n violations.push(`${relPath}:${index + 1} English template cannot require Chinese output`)\n }\n }\n }\n\n return violations\n}\n\nfunction collectLegacyPromptFiles() {\n return walk(path.join(root, 'lib', 'prompts'))\n .map((fullPath) => toRel(fullPath))\n .filter((relPath) => relPath.endsWith('.txt') && !relPath.endsWith('.zh.txt') && !relPath.endsWith('.en.txt'))\n}\n\nfunction verifyPromptCatalogCoverage() {\n const catalogPath = path.join(root, 'src', 'lib', 'prompt-i18n', 'catalog.ts')\n if (!fs.existsSync(catalogPath)) {\n fail('Missing prompt catalog file', ['src/lib/prompt-i18n/catalog.ts'])\n }\n\n const catalogText = fs.readFileSync(catalogPath, 'utf8')\n const stems = Array.from(catalogText.matchAll(/pathStem:\\s*'([^']+)'/g)).map((match) => match[1])\n if (stems.length === 0) {\n fail('No prompt pathStem found in catalog.ts')\n }\n\n const missing = []\n for (const stem of stems) {\n const zhPath = path.join(root, 'lib', 'prompts', `${stem}.zh.txt`)\n const enPath = path.join(root, 'lib', 'prompts', `${stem}.en.txt`)\n if (!fs.existsSync(zhPath)) {\n missing.push(`missing zh template: lib/prompts/${stem}.zh.txt`)\n }\n if (!fs.existsSync(enPath)) {\n missing.push(`missing en template: lib/prompts/${stem}.en.txt`)\n }\n }\n\n if (missing.length > 0) {\n fail('Prompt template coverage check failed', missing)\n }\n}\n\nconst legacyPromptFiles = collectLegacyPromptFiles()\nif (legacyPromptFiles.length > 0) {\n fail('Legacy prompt files found (.txt without locale suffix)', legacyPromptFiles)\n}\n\nverifyPromptCatalogCoverage()\n\nconst promptReadViolations = collectDirectPromptReadViolations()\nif (promptReadViolations.length > 0) {\n fail('Found direct prompt template reads', promptReadViolations)\n}\n\nconst languageViolations = collectLanguageDirectiveViolations()\nif (languageViolations.length > 0) {\n fail('Found hardcoded language directives', languageViolations)\n}\n\nconsole.log('[prompt-i18n-guard] OK')\n"
},
{
"path": "scripts/guards/prompt-json-canary-guard.mjs",
"content": "#!/usr/bin/env node\n\nimport fs from 'fs'\nimport path from 'path'\nimport process from 'process'\n\nconst root = process.cwd()\n\nconst CANARY_FILES = {\n clips: 'standards/prompt-canary/story_to_script_clips.canary.json',\n screenplay: 'standards/prompt-canary/screenplay_conversion.canary.json',\n storyboardPanels: 'standards/prompt-canary/storyboard_panels.canary.json',\n voiceAnalysis: 'standards/prompt-canary/voice_analysis.canary.json',\n}\n\nconst TEMPLATE_TOKEN_REQUIREMENTS = {\n 'novel-promotion/agent_clip': ['start', 'end', 'summary', 'location', 'characters'],\n 'novel-promotion/screenplay_conversion': [\n 'clip_id',\n 'original_text',\n 'scenes',\n 'heading',\n 'content',\n 'type',\n 'action',\n 'dialogue',\n 'voiceover',\n ],\n 'novel-promotion/agent_storyboard_plan': [\n 'panel_number',\n 'description',\n 'characters',\n 'location',\n 'scene_type',\n 'source_text',\n ],\n 'novel-promotion/agent_storyboard_detail': [\n 'panel_number',\n 'description',\n 'characters',\n 'location',\n 'scene_type',\n 'source_text',\n 'shot_type',\n 'camera_move',\n 'video_prompt',\n ],\n 'novel-promotion/agent_storyboard_insert': [\n 'panel_number',\n 'description',\n 'characters',\n 'location',\n 'scene_type',\n 'source_text',\n 'shot_type',\n 'camera_move',\n 'video_prompt',\n ],\n 'novel-promotion/voice_analysis': [\n 'lineIndex',\n 'speaker',\n 'content',\n 'emotionStrength',\n 'matchedPanel',\n 'storyboardId',\n 'panelIndex',\n ],\n}\n\nfunction fail(title, details = []) {\n console.error(`\\n[prompt-json-canary-guard] ${title}`)\n for (const line of details) {\n console.error(` - ${line}`)\n }\n process.exit(1)\n}\n\nfunction isRecord(value) {\n return !!value && typeof value === 'object' && !Array.isArray(value)\n}\n\nfunction isString(value) {\n return typeof value === 'string'\n}\n\nfunction isNumber(value) {\n return typeof value === 'number' && Number.isFinite(value)\n}\n\nfunction readJson(relativePath) {\n const fullPath = path.join(root, relativePath)\n if (!fs.existsSync(fullPath)) {\n fail('Missing canary fixture', [relativePath])\n }\n try {\n return JSON.parse(fs.readFileSync(fullPath, 'utf8'))\n } catch (error) {\n fail('Invalid canary fixture JSON', [`${relativePath}: ${error instanceof Error ? error.message : String(error)}`])\n }\n}\n\nfunction validateClipCanary(value) {\n if (!Array.isArray(value) || value.length === 0) return 'clips fixture must be a non-empty array'\n for (let i = 0; i < value.length; i += 1) {\n const row = value[i]\n if (!isRecord(row)) return `clips[${i}] must be an object`\n if (!isString(row.start) || row.start.length < 5) return `clips[${i}].start must be string length >= 5`\n if (!isString(row.end) || row.end.length < 5) return `clips[${i}].end must be string length >= 5`\n if (!isString(row.summary) || row.summary.length === 0) return `clips[${i}].summary must be non-empty string`\n if (!(row.location === null || isString(row.location))) return `clips[${i}].location must be string or null`\n if (!Array.isArray(row.characters) || !row.characters.every((item) => isString(item))) {\n return `clips[${i}].characters must be string array`\n }\n }\n return null\n}\n\nfunction validateScreenplayCanary(value) {\n if (!isRecord(value)) return 'screenplay fixture must be an object'\n if (!isString(value.clip_id) || !value.clip_id) return 'screenplay.clip_id must be non-empty string'\n if (!isString(value.original_text)) return 'screenplay.original_text must be string'\n if (!Array.isArray(value.scenes) || value.scenes.length === 0) return 'screenplay.scenes must be non-empty array'\n\n for (let i = 0; i < value.scenes.length; i += 1) {\n const scene = value.scenes[i]\n if (!isRecord(scene)) return `screenplay.scenes[${i}] must be object`\n if (!isNumber(scene.scene_number)) return `screenplay.scenes[${i}].scene_number must be number`\n if (!isRecord(scene.heading)) return `screenplay.scenes[${i}].heading must be object`\n if (!isString(scene.heading.int_ext)) return `screenplay.scenes[${i}].heading.int_ext must be string`\n if (!isString(scene.heading.location)) return `screenplay.scenes[${i}].heading.location must be string`\n if (!isString(scene.heading.time)) return `screenplay.scenes[${i}].heading.time must be string`\n if (!isString(scene.description)) return `screenplay.scenes[${i}].description must be string`\n if (!Array.isArray(scene.characters) || !scene.characters.every((item) => isString(item))) {\n return `screenplay.scenes[${i}].characters must be string array`\n }\n if (!Array.isArray(scene.content) || scene.content.length === 0) return `screenplay.scenes[${i}].content must be non-empty array`\n\n for (let j = 0; j < scene.content.length; j += 1) {\n const segment = scene.content[j]\n if (!isRecord(segment)) return `screenplay.scenes[${i}].content[${j}] must be object`\n if (!isString(segment.type)) return `screenplay.scenes[${i}].content[${j}].type must be string`\n if (segment.type === 'action') {\n if (!isString(segment.text)) return `screenplay action[${i}:${j}].text must be string`\n } else if (segment.type === 'dialogue') {\n if (!isString(segment.character)) return `screenplay dialogue[${i}:${j}].character must be string`\n if (!isString(segment.lines)) return `screenplay dialogue[${i}:${j}].lines must be string`\n if (segment.parenthetical !== undefined && !isString(segment.parenthetical)) {\n return `screenplay dialogue[${i}:${j}].parenthetical must be string when present`\n }\n } else if (segment.type === 'voiceover') {\n if (!isString(segment.text)) return `screenplay voiceover[${i}:${j}].text must be string`\n if (segment.character !== undefined && !isString(segment.character)) {\n return `screenplay voiceover[${i}:${j}].character must be string when present`\n }\n } else {\n return `screenplay.scenes[${i}].content[${j}].type must be action/dialogue/voiceover`\n }\n }\n }\n\n return null\n}\n\nfunction validateStoryboardPanelsCanary(value) {\n if (!Array.isArray(value) || value.length === 0) return 'storyboard panels fixture must be non-empty array'\n for (let i = 0; i < value.length; i += 1) {\n const panel = value[i]\n if (!isRecord(panel)) return `storyboardPanels[${i}] must be object`\n if (!isNumber(panel.panel_number)) return `storyboardPanels[${i}].panel_number must be number`\n if (!isString(panel.description)) return `storyboardPanels[${i}].description must be string`\n if (!isString(panel.location)) return `storyboardPanels[${i}].location must be string`\n if (!isString(panel.scene_type)) return `storyboardPanels[${i}].scene_type must be string`\n if (!isString(panel.source_text)) return `storyboardPanels[${i}].source_text must be string`\n if (!isString(panel.shot_type)) return `storyboardPanels[${i}].shot_type must be string`\n if (!isString(panel.camera_move)) return `storyboardPanels[${i}].camera_move must be string`\n if (!isString(panel.video_prompt)) return `storyboardPanels[${i}].video_prompt must be string`\n if (panel.duration !== undefined && !isNumber(panel.duration)) return `storyboardPanels[${i}].duration must be number when present`\n if (!Array.isArray(panel.characters)) return `storyboardPanels[${i}].characters must be array`\n for (let j = 0; j < panel.characters.length; j += 1) {\n const character = panel.characters[j]\n if (!isRecord(character)) return `storyboardPanels[${i}].characters[${j}] must be object`\n if (!isString(character.name)) return `storyboardPanels[${i}].characters[${j}].name must be string`\n if (character.appearance !== undefined && !isString(character.appearance)) {\n return `storyboardPanels[${i}].characters[${j}].appearance must be string when present`\n }\n }\n }\n return null\n}\n\nfunction validateVoiceAnalysisCanary(value) {\n if (!Array.isArray(value) || value.length === 0) return 'voice analysis fixture must be non-empty array'\n for (let i = 0; i < value.length; i += 1) {\n const row = value[i]\n if (!isRecord(row)) return `voiceAnalysis[${i}] must be object`\n if (!isNumber(row.lineIndex)) return `voiceAnalysis[${i}].lineIndex must be number`\n if (!isString(row.speaker)) return `voiceAnalysis[${i}].speaker must be string`\n if (!isString(row.content)) return `voiceAnalysis[${i}].content must be string`\n if (!isNumber(row.emotionStrength)) return `voiceAnalysis[${i}].emotionStrength must be number`\n if (row.matchedPanel !== null) {\n if (!isRecord(row.matchedPanel)) return `voiceAnalysis[${i}].matchedPanel must be object or null`\n if (!isString(row.matchedPanel.storyboardId)) return `voiceAnalysis[${i}].matchedPanel.storyboardId must be string`\n if (!isNumber(row.matchedPanel.panelIndex)) return `voiceAnalysis[${i}].matchedPanel.panelIndex must be number`\n }\n }\n return null\n}\n\nfunction checkTemplateTokens(pathStem, requiredTokens) {\n const violations = []\n for (const locale of ['zh', 'en']) {\n const relPath = `lib/prompts/${pathStem}.${locale}.txt`\n const fullPath = path.join(root, relPath)\n if (!fs.existsSync(fullPath)) {\n violations.push(`missing template: ${relPath}`)\n continue\n }\n const content = fs.readFileSync(fullPath, 'utf8')\n for (const token of requiredTokens) {\n if (!content.includes(token)) {\n violations.push(`missing token ${token} in ${relPath}`)\n }\n }\n }\n return violations\n}\n\nconst violations = []\n\nconst clipsErr = validateClipCanary(readJson(CANARY_FILES.clips))\nif (clipsErr) violations.push(clipsErr)\n\nconst screenplayErr = validateScreenplayCanary(readJson(CANARY_FILES.screenplay))\nif (screenplayErr) violations.push(screenplayErr)\n\nconst panelsErr = validateStoryboardPanelsCanary(readJson(CANARY_FILES.storyboardPanels))\nif (panelsErr) violations.push(panelsErr)\n\nconst voiceErr = validateVoiceAnalysisCanary(readJson(CANARY_FILES.voiceAnalysis))\nif (voiceErr) violations.push(voiceErr)\n\nfor (const [pathStem, requiredTokens] of Object.entries(TEMPLATE_TOKEN_REQUIREMENTS)) {\n violations.push(...checkTemplateTokens(pathStem, requiredTokens))\n}\n\nif (violations.length > 0) {\n fail('JSON schema canary check failed', violations)\n}\n\nconsole.log('[prompt-json-canary-guard] OK')\n"
},
{
"path": "scripts/guards/prompt-semantic-regression.mjs",
"content": "#!/usr/bin/env node\n\nimport fs from 'fs'\nimport path from 'path'\nimport process from 'process'\n\nconst root = process.cwd()\nconst catalogPath = path.join(root, 'src', 'lib', 'prompt-i18n', 'catalog.ts')\nconst chineseCharPattern = /[\\p{Script=Han}]/u\nconst singlePlaceholderPattern = /\\{([A-Za-z0-9_]+)\\}/g\nconst doublePlaceholderPattern = /\\{\\{([A-Za-z0-9_]+)\\}\\}/g\n\nconst criticalTemplateTokens = new Map([\n ['novel-promotion/voice_analysis', ['\"lineIndex\"', '\"speaker\"', '\"content\"', '\"emotionStrength\"', '\"matchedPanel\"']],\n ['novel-promotion/agent_storyboard_plan', ['\"panel_number\"', '\"description\"', '\"characters\"', '\"location\"', '\"scene_type\"', '\"source_text\"', '\"shot_type\"', '\"camera_move\"', '\"video_prompt\"']],\n ['novel-promotion/agent_storyboard_detail', ['\"panel_number\"', '\"description\"', '\"characters\"', '\"location\"', '\"scene_type\"', '\"source_text\"', '\"shot_type\"', '\"camera_move\"', '\"video_prompt\"']],\n ['novel-promotion/agent_storyboard_insert', ['\"panel_number\"', '\"description\"', '\"characters\"', '\"location\"', '\"scene_type\"', '\"source_text\"', '\"shot_type\"', '\"camera_move\"', '\"video_prompt\"']],\n ['novel-promotion/screenplay_conversion', ['\"clip_id\"', '\"scenes\"', '\"heading\"', '\"content\"', '\"dialogue\"', '\"voiceover\"']],\n ['novel-promotion/select_location', ['\"locations\"', '\"name\"', '\"summary\"', '\"descriptions\"']],\n ['novel-promotion/episode_split', ['\"analysis\"', '\"episodes\"', '\"startMarker\"', '\"endMarker\"', '\"validation\"']],\n ['novel-promotion/image_prompt_modify', ['\"image_prompt\"', '\"video_prompt\"']],\n ['novel-promotion/character_create', ['\"prompt\"']],\n ['novel-promotion/location_create', ['\"prompt\"']],\n])\n\nfunction fail(title, details = []) {\n console.error(`\\n[prompt-semantic-regression] ${title}`)\n for (const line of details) {\n console.error(` - ${line}`)\n }\n process.exit(1)\n}\n\nfunction parseCatalog(text) {\n const entries = []\n const entryPattern = /pathStem:\\s*'([^']+)'\\s*,[\\s\\S]*?variableKeys:\\s*\\[([\\s\\S]*?)\\]\\s*,/g\n for (const match of text.matchAll(entryPattern)) {\n const pathStem = match[1]\n const rawKeys = match[2] || ''\n const keys = Array.from(rawKeys.matchAll(/'([^']+)'/g)).map((item) => item[1])\n entries.push({ pathStem, variableKeys: keys })\n }\n return entries\n}\n\nfunction extractPlaceholders(template) {\n const keys = new Set()\n for (const match of template.matchAll(singlePlaceholderPattern)) {\n if (match[1]) keys.add(match[1])\n }\n for (const match of template.matchAll(doublePlaceholderPattern)) {\n if (match[1]) keys.add(match[1])\n }\n return Array.from(keys)\n}\n\nif (!fs.existsSync(catalogPath)) {\n fail('catalog.ts not found', ['src/lib/prompt-i18n/catalog.ts'])\n}\n\nconst catalogText = fs.readFileSync(catalogPath, 'utf8')\nconst entries = parseCatalog(catalogText)\nif (entries.length === 0) {\n fail('failed to parse prompt catalog entries')\n}\n\nconst violations = []\nfor (const entry of entries) {\n const templatePath = path.join(root, 'lib', 'prompts', `${entry.pathStem}.en.txt`)\n if (!fs.existsSync(templatePath)) {\n violations.push(`missing template: lib/prompts/${entry.pathStem}.en.txt`)\n continue\n }\n\n const template = fs.readFileSync(templatePath, 'utf8')\n if (chineseCharPattern.test(template)) {\n violations.push(`unexpected Chinese content in English template: lib/prompts/${entry.pathStem}.en.txt`)\n }\n\n const placeholders = extractPlaceholders(template)\n const placeholderSet = new Set(placeholders)\n const variableKeySet = new Set(entry.variableKeys)\n\n for (const key of entry.variableKeys) {\n if (!placeholderSet.has(key)) {\n violations.push(`missing placeholder {${key}} in lib/prompts/${entry.pathStem}.en.txt`)\n }\n }\n\n for (const key of placeholders) {\n if (!variableKeySet.has(key)) {\n violations.push(`unexpected placeholder {${key}} in lib/prompts/${entry.pathStem}.en.txt`)\n }\n }\n\n const requiredTokens = criticalTemplateTokens.get(entry.pathStem) || []\n for (const token of requiredTokens) {\n if (!template.includes(token)) {\n violations.push(`missing semantic token ${token} in lib/prompts/${entry.pathStem}.en.txt`)\n }\n }\n}\n\nif (violations.length > 0) {\n fail('semantic regression check failed', violations)\n}\n\nconsole.log(`[prompt-semantic-regression] OK (${entries.length} templates checked)`)\n"
},
{
"path": "scripts/guards/task-loading-baseline.json",
"content": "{\n \"allowedDirectTaskStateUsageFiles\": [\n \"src/lib/query/hooks/useTaskTargetStates.ts\",\n \"src/lib/query/hooks/useTaskPresentation.ts\",\n \"src/lib/query/hooks/useProjectAssets.ts\",\n \"src/lib/query/hooks/useGlobalAssets.ts\"\n ],\n \"allowedLegacyGeneratingUsageFiles\": []\n}\n"
},
{
"path": "scripts/guards/task-loading-guard.mjs",
"content": "#!/usr/bin/env node\n\nimport fs from 'fs'\nimport path from 'path'\nimport process from 'process'\n\nconst workspaceRoot = process.cwd()\nconst baselinePath = path.join(workspaceRoot, 'scripts/guards/task-loading-baseline.json')\n\nfunction walkFiles(dir, out = []) {\n const entries = fs.readdirSync(dir, { withFileTypes: true })\n for (const entry of entries) {\n if (entry.name === 'node_modules' || entry.name === '.git' || entry.name === '.next') continue\n const fullPath = path.join(dir, entry.name)\n if (entry.isDirectory()) {\n walkFiles(fullPath, out)\n } else {\n out.push(fullPath)\n }\n }\n return out\n}\n\nfunction toPosixRelative(filePath) {\n return path.relative(workspaceRoot, filePath).split(path.sep).join('/')\n}\n\nfunction collectMatches(files, pattern) {\n const matches = []\n for (const fullPath of files) {\n if (!fullPath.endsWith('.ts') && !fullPath.endsWith('.tsx')) continue\n const relPath = toPosixRelative(fullPath)\n const content = fs.readFileSync(fullPath, 'utf8')\n const lines = content.split('\\n')\n for (let i = 0; i < lines.length; i += 1) {\n if (lines[i].includes(pattern)) {\n matches.push(`${relPath}:${i + 1}`)\n }\n }\n }\n return matches\n}\n\nfunction fail(title, lines) {\n console.error(`\\n[task-loading-guard] ${title}`)\n for (const line of lines) {\n console.error(` - ${line}`)\n }\n process.exit(1)\n}\n\nif (!fs.existsSync(baselinePath)) {\n fail('Missing baseline file', [toPosixRelative(baselinePath)])\n}\n\nconst baseline = JSON.parse(fs.readFileSync(baselinePath, 'utf8'))\nconst allowedFiles = new Set(baseline.allowedDirectTaskStateUsageFiles || [])\nconst allowedLegacyGeneratingFiles = new Set(baseline.allowedLegacyGeneratingUsageFiles || [])\nconst allFiles = walkFiles(path.join(workspaceRoot, 'src'))\n\nconst directTaskStateUsage = collectMatches(allFiles, 'useTaskTargetStates(')\nconst directUsageOutOfAllowlist = directTaskStateUsage\n .map((entry) => entry.split(':')[0])\n .filter((file) => !allowedFiles.has(file))\n\nif (directUsageOutOfAllowlist.length > 0) {\n fail(\n 'Found component-level direct useTaskTargetStates outside baseline allowlist',\n Array.from(new Set(directUsageOutOfAllowlist)),\n )\n}\n\nconst crossDomainLabels = collectMatches(allFiles, 'video.panelCard.generating')\nif (crossDomainLabels.length > 0) {\n fail('Found cross-domain loading label reuse (video.panelCard.generating)', crossDomainLabels)\n}\n\nconst uiFiles = allFiles.filter((file) => {\n const relPath = toPosixRelative(file)\n return relPath.startsWith('src/app/') || relPath.startsWith('src/components/')\n})\nconst legacyGeneratingPatterns = [\n 'appearance.generating',\n 'panel.generatingImage',\n 'shot.generatingImage',\n 'line.generating',\n]\nconst legacyGeneratingMatches = legacyGeneratingPatterns.flatMap((pattern) =>\n collectMatches(uiFiles, pattern),\n)\nconst legacyGeneratingOutOfAllowlist = legacyGeneratingMatches\n .map((entry) => entry.split(':')[0])\n .filter((file) => !allowedLegacyGeneratingFiles.has(file))\nif (legacyGeneratingOutOfAllowlist.length > 0) {\n fail(\n 'Found legacy generating truth usage in UI components',\n Array.from(new Set(legacyGeneratingOutOfAllowlist)),\n )\n}\n\nconst hooksIndexPath = path.join(workspaceRoot, 'src/lib/query/hooks/index.ts')\nif (fs.existsSync(hooksIndexPath)) {\n const hooksIndex = fs.readFileSync(hooksIndexPath, 'utf8')\n const bannedReexports = [\n {\n pattern: /export\\s*\\{[^}]*useGenerateCharacterImage[^}]*\\}\\s*from\\s*['\"]\\.\\/useGlobalAssets['\"]/m,\n message: 'hooks/index.ts must not export useGenerateCharacterImage from useGlobalAssets',\n },\n {\n pattern: /export\\s*\\{[^}]*useGenerateLocationImage[^}]*\\}\\s*from\\s*['\"]\\.\\/useGlobalAssets['\"]/m,\n message: 'hooks/index.ts must not export useGenerateLocationImage from useGlobalAssets',\n },\n {\n pattern: /export\\s*\\{[^}]*useGenerateProjectCharacterImage[^}]*\\}\\s*from\\s*['\"]\\.\\/useProjectAssets['\"]/m,\n message: 'hooks/index.ts must not export useGenerateProjectCharacterImage from useProjectAssets',\n },\n {\n pattern: /export\\s*\\{[^}]*useGenerateProjectLocationImage[^}]*\\}\\s*from\\s*['\"]\\.\\/useProjectAssets['\"]/m,\n message: 'hooks/index.ts must not export useGenerateProjectLocationImage from useProjectAssets',\n },\n ]\n\n const violations = bannedReexports\n .filter((item) => item.pattern.test(hooksIndex))\n .map((item) => item.message)\n\n if (violations.length > 0) {\n fail('Found non-canonical mutation re-exports', violations)\n }\n}\n\nconsole.log('[task-loading-guard] OK')\n"
},
{
"path": "scripts/guards/task-state-unification-guard.sh",
"content": "#!/usr/bin/env bash\nset -euo pipefail\n\nfailed=0\n\ncheck_absent() {\n local label=\"$1\"\n local pattern=\"$2\"\n shift 2\n local output\n output=\"$(git grep --untracked -nE \"$pattern\" -- \"$@\" || true)\"\n if [[ -n \"$output\" ]]; then\n echo \"$output\"\n echo \"::error title=${label}::${label}\"\n failed=1\n fi\n}\n\ncheck_absent \\\n \"Do not branch UI status on cancelled\" \\\n \"status[[:space:]]*===[[:space:]]*['\\\\\\\"]cancelled['\\\\\\\"]|status[[:space:]]*==[[:space:]]*['\\\\\\\"]cancelled['\\\\\\\"]\" \\\n src/app \\\n src/components \\\n src/features \\\n src/lib/query\n\ncheck_absent \\\n \"useTaskHandoff is forbidden\" \\\n \"useTaskHandoff\" \\\n src\n\ncheck_absent \\\n \"Do not use legacy task hooks in app layer\" \\\n \"useActiveTasks\\\\(|useTaskStatus\\\\(\" \\\n src/app \\\n src/features\n\nif [[ \"$failed\" -ne 0 ]]; then\n exit 1\nfi\n\necho \"task-state-unification guard passed\"\n"
},
{
"path": "scripts/guards/task-status-cutover-audit.sh",
"content": "#!/usr/bin/env bash\nset -euo pipefail\n\nROOT_DIR=\"$(git rev-parse --show-toplevel)\"\ncd \"$ROOT_DIR\"\n\nFAILED=0\n\nprint_header() {\n echo\n echo \"============================================================\"\n echo \"$1\"\n echo \"============================================================\"\n}\n\nprint_ok() {\n echo \"[PASS] $1\"\n}\n\nprint_fail() {\n echo \"[FAIL] $1\"\n}\n\nrun_zero_match_check() {\n local title=\"$1\"\n local pattern=\"$2\"\n shift 2\n local paths=(\"$@\")\n local output\n output=\"$(git grep -n -E \"$pattern\" -- \"${paths[@]}\" || true)\"\n if [[ -z \"$output\" ]]; then\n print_ok \"$title\"\n else\n print_fail \"$title\"\n echo \"$output\"\n FAILED=1\n fi\n}\n\nrun_usetasktargetstates_check() {\n local title=\"useTaskTargetStates 仅允许在 useProjectAssets/useGlobalAssets 中使用\"\n local output\n output=\"$(git grep -n \"useTaskTargetStates\" -- src || true)\"\n\n if [[ -z \"$output\" ]]; then\n print_ok \"$title (当前 0 命中)\"\n return\n fi\n\n local filtered\n filtered=\"$(echo \"$output\" | grep -v \"src/lib/query/hooks/useProjectAssets.ts\" | grep -v \"src/lib/query/hooks/useGlobalAssets.ts\" || true)\"\n\n if [[ -z \"$filtered\" ]]; then\n print_ok \"$title\"\n else\n print_fail \"$title\"\n echo \"$filtered\"\n FAILED=1\n fi\n}\n\nprint_header \"Task Status Cutover Audit\"\n\nrun_zero_match_check \\\n \"禁止 useTaskHandoff\" \\\n \"useTaskHandoff\" \\\n src\n\nrun_zero_match_check \\\n \"禁止 manualRegeneratingItems/setRegeneratingItems/clearRegeneratingItem\" \\\n \"manualRegeneratingItems|setRegeneratingItems|clearRegeneratingItem\" \\\n src\n\nrun_zero_match_check \\\n \"禁止业务层直接判断 status ===/!== cancelled\" \\\n \"status\\\\s*===\\\\s*['\\\\\\\"]cancelled['\\\\\\\"]|status\\\\s*!==\\\\s*['\\\\\\\"]cancelled['\\\\\\\"]\" \\\n src\n\nrun_zero_match_check \\\n \"禁止 generatingImage/generatingVideo/generatingLipSync 字段\" \\\n \"\\\\bgeneratingImage\\\\b|\\\\bgeneratingVideo\\\\b|\\\\bgeneratingLipSync\\\\b\" \\\n src\n\nrun_usetasktargetstates_check\n\nrun_zero_match_check \\\n \"禁止 novel-promotion/asset-hub/shared-assets 中 useState(false) 作为生成态命名\" \\\n \"const \\\\[[^\\\\]]*(Generating|Regenerating|WaitingForGeneration|AnalyzingAssets|GeneratingAll|CopyingFromGlobal)[^\\\\]]*\\\\]\\\\s*=\\\\s*useState\\\\(false\\\\)\" \\\n \"src/app/[locale]/workspace/[projectId]/modes/novel-promotion\" \\\n \"src/app/[locale]/workspace/asset-hub\" \\\n \"src/components/shared/assets\"\n\nprint_header \"Audit Result\"\nif [[ \"$FAILED\" -eq 0 ]]; then\n echo \"All checks passed.\"\n exit 0\nfi\n\necho \"Audit failed. Please fix findings above.\"\nexit 1\n"
},
{
"path": "scripts/guards/task-submit-compensation-guard.mjs",
"content": "#!/usr/bin/env node\n\nimport fs from 'fs'\nimport path from 'path'\nimport process from 'process'\nimport { pathToFileURL } from 'url'\n\nconst root = process.cwd()\nconst apiDir = path.join(root, 'src', 'app', 'api')\nconst CREATE_PATTERN = /\\.\\s*create\\s*\\(/\nconst SUBMIT_TASK_PATTERN = /\\bsubmitTask\\s*\\(/\nconst ROLLBACK_PATTERN = /rollback|compensat/i\n\nfunction fail(title, details = []) {\n process.stderr.write(`\\n[task-submit-compensation-guard] ${title}\\n`)\n for (const detail of details) {\n process.stderr.write(` - ${detail}\\n`)\n }\n process.exit(1)\n}\n\nfunction walk(dir, out = []) {\n if (!fs.existsSync(dir)) return out\n const entries = fs.readdirSync(dir, { withFileTypes: true })\n for (const entry of entries) {\n if (entry.name === '.git' || entry.name === '.next' || entry.name === 'node_modules') continue\n const fullPath = path.join(dir, entry.name)\n if (entry.isDirectory()) {\n walk(fullPath, out)\n continue\n }\n if (entry.name === 'route.ts') out.push(fullPath)\n }\n return out\n}\n\nfunction toRel(fullPath) {\n return path.relative(root, fullPath).split(path.sep).join('/')\n}\n\nexport function inspectTaskSubmitCompensation(relPath, content) {\n if (!CREATE_PATTERN.test(content)) return []\n if (!SUBMIT_TASK_PATTERN.test(content)) return []\n if (ROLLBACK_PATTERN.test(content)) return []\n return [\n `${relPath} creates data before submitTask without explicit rollback/compensation marker`,\n ]\n}\n\nexport function findTaskSubmitCompensationViolations(scanRoot = root) {\n const routesRoot = path.join(scanRoot, 'src', 'app', 'api')\n return walk(routesRoot)\n .map((fullPath) => {\n const relPath = path.relative(scanRoot, fullPath).split(path.sep).join('/')\n const content = fs.readFileSync(fullPath, 'utf8')\n return inspectTaskSubmitCompensation(relPath, content)\n })\n .flat()\n}\n\nexport function main() {\n if (!fs.existsSync(apiDir)) {\n fail('Missing src/app/api directory')\n }\n\n const routeFiles = walk(apiDir)\n const violations = routeFiles\n .map((fullPath) => {\n const relPath = toRel(fullPath)\n const content = fs.readFileSync(fullPath, 'utf8')\n return inspectTaskSubmitCompensation(relPath, content)\n })\n .flat()\n\n if (violations.length > 0) {\n fail('Found create+submitTask routes without compensation marker', violations)\n }\n\n process.stdout.write(`[task-submit-compensation-guard] OK routes=${routeFiles.length}\\n`)\n}\n\nif (process.argv[1] && import.meta.url === pathToFileURL(process.argv[1]).href) {\n main()\n}\n"
},
{
"path": "scripts/guards/task-target-states-no-polling-guard.mjs",
"content": "#!/usr/bin/env node\n\nimport fs from 'fs'\nimport path from 'path'\nimport process from 'process'\n\nconst root = process.cwd()\n\nfunction fail(title, details = []) {\n console.error(`\\n[task-target-states-no-polling-guard] ${title}`)\n for (const line of details) {\n console.error(` - ${line}`)\n }\n process.exit(1)\n}\n\nfunction readFile(relativePath) {\n const fullPath = path.join(root, relativePath)\n if (!fs.existsSync(fullPath)) {\n fail('Missing required file', [relativePath])\n }\n return fs.readFileSync(fullPath, 'utf8')\n}\n\nfunction walk(dir, out = []) {\n const entries = fs.readdirSync(dir, { withFileTypes: true })\n for (const entry of entries) {\n if (entry.name === '.git' || entry.name === '.next' || entry.name === 'node_modules') continue\n const full = path.join(dir, entry.name)\n if (entry.isDirectory()) {\n walk(full, out)\n } else {\n out.push(full)\n }\n }\n return out\n}\n\nfunction toRel(fullPath) {\n return path.relative(root, fullPath).split(path.sep).join('/')\n}\n\nfunction collectPattern(pattern) {\n const files = walk(path.join(root, 'src'))\n const hits = []\n for (const fullPath of files) {\n if (!fullPath.endsWith('.ts') && !fullPath.endsWith('.tsx')) continue\n const text = fs.readFileSync(fullPath, 'utf8')\n const lines = text.split('\\n')\n for (let i = 0; i < lines.length; i += 1) {\n if (pattern.test(lines[i])) {\n hits.push(`${toRel(fullPath)}:${i + 1}`)\n }\n }\n }\n return hits\n}\n\nconst refetchIntervalMsHits = collectPattern(/\\brefetchIntervalMs\\b/)\nif (refetchIntervalMsHits.length > 0) {\n fail('Found forbidden refetchIntervalMs usage', refetchIntervalMsHits)\n}\n\nconst voiceStagePath =\n 'src/app/[locale]/workspace/[projectId]/modes/novel-promotion/components/VoiceStage.tsx'\nconst voiceStageText = readFile(voiceStagePath)\nif (voiceStageText.includes('setInterval(')) {\n fail('VoiceStage must not use timer polling', [voiceStagePath])\n}\n\nconst targetStateMapPath = 'src/lib/query/hooks/useTaskTargetStateMap.ts'\nconst targetStateMapText = readFile(targetStateMapPath)\nif (!/refetchInterval:\\s*false/.test(targetStateMapText)) {\n fail('useTaskTargetStateMap must keep refetchInterval disabled', [targetStateMapPath])\n}\n\nconst ssePath = 'src/lib/query/hooks/useSSE.ts'\nconst sseText = readFile(ssePath)\nconst targetStatesInvalidateExprMatch = sseText.match(\n /const shouldInvalidateTargetStates\\s*=\\s*([\\s\\S]*?)\\n\\s*\\n/,\n)\nif (!targetStatesInvalidateExprMatch) {\n fail('Unable to locate shouldInvalidateTargetStates expression', [ssePath])\n}\nconst targetStatesInvalidateExpr = targetStatesInvalidateExprMatch[1]\nif (!/TASK_EVENT_TYPE\\.COMPLETED/.test(targetStatesInvalidateExpr) || !/TASK_EVENT_TYPE\\.FAILED/.test(targetStatesInvalidateExpr)) {\n fail('useSSE must invalidate target states only for terminal events', [ssePath])\n}\nif (/TASK_EVENT_TYPE\\.CREATED/.test(targetStatesInvalidateExpr)) {\n fail('useSSE target-state invalidation must not include CREATED', [ssePath])\n}\nif (/TASK_EVENT_TYPE\\.PROCESSING/.test(targetStatesInvalidateExpr)) {\n fail('useSSE target-state invalidation must not include PROCESSING', [ssePath])\n}\n\nconsole.log('[task-target-states-no-polling-guard] OK')\n"
},
{
"path": "scripts/guards/test-behavior-quality-guard.mjs",
"content": "#!/usr/bin/env node\n\nimport fs from 'fs'\nimport path from 'path'\n\nconst root = process.cwd()\nconst targetDirs = [\n path.join(root, 'tests', 'integration', 'api', 'contract'),\n path.join(root, 'tests', 'integration', 'provider'),\n path.join(root, 'tests', 'integration', 'chain'),\n path.join(root, 'tests', 'system'),\n path.join(root, 'tests', 'regression'),\n]\n\nfunction fail(title, details = []) {\n console.error(`\\n[test-behavior-quality-guard] ${title}`)\n for (const detail of details) {\n console.error(` - ${detail}`)\n }\n process.exit(1)\n}\n\nfunction walk(dir, out = []) {\n if (!fs.existsSync(dir)) return out\n const entries = fs.readdirSync(dir, { withFileTypes: true })\n for (const entry of entries) {\n if (entry.name === '.git' || entry.name === 'node_modules') continue\n const full = path.join(dir, entry.name)\n if (entry.isDirectory()) {\n walk(full, out)\n continue\n }\n if (entry.isFile() && entry.name.endsWith('.test.ts')) out.push(full)\n }\n return out\n}\n\nfunction toRel(fullPath) {\n return path.relative(root, fullPath).split(path.sep).join('/')\n}\n\nconst files = targetDirs.flatMap((dir) => walk(dir))\nif (files.length === 0) {\n fail('No target test files found', targetDirs.map((dir) => toRel(dir)))\n}\n\nconst violations = []\n\nfor (const file of files) {\n const rel = toRel(file)\n const text = fs.readFileSync(file, 'utf8')\n\n const hasSourceRead = /(readFileSync|fs\\.readFileSync)\\s*\\([\\s\\S]{0,240}src\\/(app|lib)\\//m.test(text)\n if (hasSourceRead) {\n violations.push(`${rel}: reading source code text is forbidden in behavior contract/chain tests`)\n }\n\n const forbiddenStringContracts = [\n /toContain\\(\\s*['\"]apiHandler['\"]\\s*\\)/,\n /toContain\\(\\s*['\"]submitTask['\"]\\s*\\)/,\n /toContain\\(\\s*['\"]maybeSubmitLLMTask['\"]\\s*\\)/,\n /includes\\(\\s*['\"]apiHandler['\"]\\s*\\)/,\n /includes\\(\\s*['\"]submitTask['\"]\\s*\\)/,\n /includes\\(\\s*['\"]maybeSubmitLLMTask['\"]\\s*\\)/,\n ]\n\n for (const pattern of forbiddenStringContracts) {\n if (pattern.test(text)) {\n violations.push(`${rel}: forbidden structural string assertion matched ${pattern}`)\n break\n }\n }\n\n const hasWeakCallAssertion = /toHaveBeenCalled\\(\\s*\\)/.test(text)\n const hasStrongCallAssertion = /toHaveBeenCalledWith\\(/.test(text)\n if (hasWeakCallAssertion && !hasStrongCallAssertion) {\n violations.push(`${rel}: has toHaveBeenCalled() without any toHaveBeenCalledWith() result assertions`)\n }\n}\n\nif (violations.length > 0) {\n fail('Behavior quality violations found', violations)\n}\n\nconsole.log(`[test-behavior-quality-guard] OK files=${files.length}`)\n"
},
{
"path": "scripts/guards/test-behavior-route-coverage-guard.mjs",
"content": "#!/usr/bin/env node\n\nimport fs from 'fs'\nimport path from 'path'\n\nconst root = process.cwd()\nconst catalogPath = path.join(root, 'tests', 'contracts', 'route-catalog.ts')\nconst matrixPath = path.join(root, 'tests', 'contracts', 'route-behavior-matrix.ts')\n\nfunction fail(title, details = []) {\n console.error(`\\n[test-behavior-route-coverage-guard] ${title}`)\n for (const detail of details) {\n console.error(` - ${detail}`)\n }\n process.exit(1)\n}\n\nif (!fs.existsSync(catalogPath)) {\n fail('route catalog is missing', ['tests/contracts/route-catalog.ts'])\n}\nif (!fs.existsSync(matrixPath)) {\n fail('route behavior matrix is missing', ['tests/contracts/route-behavior-matrix.ts'])\n}\n\nconst catalogText = fs.readFileSync(catalogPath, 'utf8')\nconst matrixText = fs.readFileSync(matrixPath, 'utf8')\n\nif (!matrixText.includes('ROUTE_CATALOG.map')) {\n fail('route behavior matrix must derive entries from ROUTE_CATALOG.map')\n}\n\nconst routeFilesBlockMatch = catalogText.match(/const ROUTE_FILES = \\[([\\s\\S]*?)\\] as const/)\nif (!routeFilesBlockMatch) {\n fail('unable to parse ROUTE_FILES block from route catalog')\n}\nconst routeFilesBlock = routeFilesBlockMatch ? routeFilesBlockMatch[1] : ''\nconst routeCount = Array.from(routeFilesBlock.matchAll(/'src\\/app\\/api\\/[^']+\\/route\\.ts'/g)).length\nif (routeCount === 0) {\n fail('no routes detected in route catalog')\n}\n\nconst testFiles = Array.from(matrixText.matchAll(/'tests\\/[a-zA-Z0-9_\\-/.]+\\.test\\.ts'/g))\n .map((match) => match[0].slice(1, -1))\n\nif (testFiles.length === 0) {\n fail('route behavior matrix does not declare any behavior test files')\n}\n\nconst missingTests = Array.from(new Set(testFiles)).filter((file) => !fs.existsSync(path.join(root, file)))\nif (missingTests.length > 0) {\n fail('route behavior matrix references missing test files', missingTests)\n}\n\nconsole.log(`[test-behavior-route-coverage-guard] OK routes=${routeCount} tests=${new Set(testFiles).size}`)\n"
},
{
"path": "scripts/guards/test-behavior-tasktype-coverage-guard.mjs",
"content": "#!/usr/bin/env node\n\nimport fs from 'fs'\nimport path from 'path'\n\nconst root = process.cwd()\nconst catalogPath = path.join(root, 'tests', 'contracts', 'task-type-catalog.ts')\nconst matrixPath = path.join(root, 'tests', 'contracts', 'tasktype-behavior-matrix.ts')\n\nfunction fail(title, details = []) {\n console.error(`\\n[test-behavior-tasktype-coverage-guard] ${title}`)\n for (const detail of details) {\n console.error(` - ${detail}`)\n }\n process.exit(1)\n}\n\nif (!fs.existsSync(catalogPath)) {\n fail('task type catalog is missing', ['tests/contracts/task-type-catalog.ts'])\n}\nif (!fs.existsSync(matrixPath)) {\n fail('tasktype behavior matrix is missing', ['tests/contracts/tasktype-behavior-matrix.ts'])\n}\n\nconst catalogText = fs.readFileSync(catalogPath, 'utf8')\nconst matrixText = fs.readFileSync(matrixPath, 'utf8')\n\nif (!matrixText.includes('TASK_TYPE_CATALOG.map')) {\n fail('tasktype behavior matrix must derive entries from TASK_TYPE_CATALOG.map')\n}\n\nconst taskTypeCount = Array.from(catalogText.matchAll(/\\[TASK_TYPE\\.([A-Z_]+)\\]/g)).length\nif (taskTypeCount === 0) {\n fail('no task types detected in task type catalog')\n}\n\nconst testFiles = Array.from(matrixText.matchAll(/'tests\\/[a-zA-Z0-9_\\-/.]+\\.test\\.ts'/g))\n .map((match) => match[0].slice(1, -1))\n\nif (testFiles.length === 0) {\n fail('tasktype behavior matrix does not declare any behavior test files')\n}\n\nconst missingTests = Array.from(new Set(testFiles)).filter((file) => !fs.existsSync(path.join(root, file)))\nif (missingTests.length > 0) {\n fail('tasktype behavior matrix references missing test files', missingTests)\n}\n\nconsole.log(`[test-behavior-tasktype-coverage-guard] OK taskTypes=${taskTypeCount} tests=${new Set(testFiles).size}`)\n"
},
{
"path": "scripts/guards/test-route-coverage-guard.mjs",
"content": "#!/usr/bin/env node\n\nimport fs from 'fs'\nimport path from 'path'\n\nconst root = process.cwd()\nconst apiDir = path.join(root, 'src', 'app', 'api')\nconst catalogPath = path.join(root, 'tests', 'contracts', 'route-catalog.ts')\n\nfunction fail(title, details = []) {\n console.error(`\\n[test-route-coverage-guard] ${title}`)\n for (const detail of details) {\n console.error(` - ${detail}`)\n }\n process.exit(1)\n}\n\nfunction walk(dir, out = []) {\n if (!fs.existsSync(dir)) return out\n const entries = fs.readdirSync(dir, { withFileTypes: true })\n for (const entry of entries) {\n if (entry.name === '.git' || entry.name === '.next' || entry.name === 'node_modules') continue\n const fullPath = path.join(dir, entry.name)\n if (entry.isDirectory()) {\n walk(fullPath, out)\n continue\n }\n if (entry.name === 'route.ts') out.push(fullPath)\n }\n return out\n}\n\nfunction toRel(fullPath) {\n return path.relative(root, fullPath).split(path.sep).join('/')\n}\n\nif (!fs.existsSync(catalogPath)) {\n fail('route-catalog.ts is missing', ['tests/contracts/route-catalog.ts'])\n}\n\nconst actualRoutes = walk(apiDir).map(toRel).sort()\nconst catalogText = fs.readFileSync(catalogPath, 'utf8')\nconst catalogRoutes = Array.from(catalogText.matchAll(/'src\\/app\\/api\\/[^']+\\/route\\.ts'/g))\n .map((match) => match[0].slice(1, -1))\n .sort()\n\nconst missingInCatalog = actualRoutes.filter((routeFile) => !catalogRoutes.includes(routeFile))\nconst staleInCatalog = catalogRoutes.filter((routeFile) => !actualRoutes.includes(routeFile))\n\nif (missingInCatalog.length > 0) {\n fail('Missing routes in tests/contracts/route-catalog.ts', missingInCatalog)\n}\nif (staleInCatalog.length > 0) {\n fail('Stale route entries found in tests/contracts/route-catalog.ts', staleInCatalog)\n}\n\nconsole.log(`[test-route-coverage-guard] OK routes=${actualRoutes.length}`)\n"
},
{
"path": "scripts/guards/test-tasktype-coverage-guard.mjs",
"content": "#!/usr/bin/env node\n\nimport fs from 'fs'\nimport path from 'path'\n\nconst root = process.cwd()\nconst taskTypesPath = path.join(root, 'src', 'lib', 'task', 'types.ts')\nconst catalogPath = path.join(root, 'tests', 'contracts', 'task-type-catalog.ts')\n\nfunction fail(title, details = []) {\n console.error(`\\n[test-tasktype-coverage-guard] ${title}`)\n for (const detail of details) {\n console.error(` - ${detail}`)\n }\n process.exit(1)\n}\n\nif (!fs.existsSync(taskTypesPath)) {\n fail('Task type source file is missing', ['src/lib/task/types.ts'])\n}\nif (!fs.existsSync(catalogPath)) {\n fail('Task type catalog file is missing', ['tests/contracts/task-type-catalog.ts'])\n}\n\nconst taskTypesText = fs.readFileSync(taskTypesPath, 'utf8')\nconst catalogText = fs.readFileSync(catalogPath, 'utf8')\n\nconst taskTypeBlockMatch = taskTypesText.match(/export const TASK_TYPE = \\{([\\s\\S]*?)\\n\\} as const/)\nif (!taskTypeBlockMatch) {\n fail('Unable to parse TASK_TYPE block from src/lib/task/types.ts')\n}\nconst taskTypeBlock = taskTypeBlockMatch ? taskTypeBlockMatch[1] : ''\nconst taskTypeKeys = Array.from(taskTypeBlock.matchAll(/^\\s+([A-Z_]+):\\s'[^']+',?$/gm)).map((match) => match[1])\nconst catalogKeys = Array.from(catalogText.matchAll(/\\[TASK_TYPE\\.([A-Z_]+)\\]/g)).map((match) => match[1])\n\nconst missingKeys = taskTypeKeys.filter((key) => !catalogKeys.includes(key))\nconst staleKeys = catalogKeys.filter((key) => !taskTypeKeys.includes(key))\n\nif (missingKeys.length > 0) {\n fail('Missing TASK_TYPE owners in tests/contracts/task-type-catalog.ts', missingKeys)\n}\nif (staleKeys.length > 0) {\n fail('Stale TASK_TYPE keys in tests/contracts/task-type-catalog.ts', staleKeys)\n}\n\nconsole.log(`[test-tasktype-coverage-guard] OK taskTypes=${taskTypeKeys.length}`)\n"
},
{
"path": "scripts/media-archive-legacy-refs.ts",
"content": "import { logInfo as _ulogInfo, logError as _ulogError } from '@/lib/logging/core'\nimport { createHash } from 'node:crypto'\nimport { promises as fs } from 'node:fs'\nimport path from 'node:path'\nimport { prisma } from '@/lib/prisma'\nimport { MEDIA_MODEL_MAPPINGS } from './media-mapping'\n\nconst BACKUP_ROOT = path.join(process.cwd(), 'data', 'migration-backups')\nconst BATCH_SIZE = 500\ntype DynamicModel = {\n findMany: (args: unknown) => Promise>>\n createMany?: (args: unknown) => Promise\n}\nconst prismaDynamic = prisma as unknown as Record\n\nfunction nowStamp() {\n return new Date().toISOString().replace(/[:.]/g, '-')\n}\n\nfunction checksum(value: string) {\n return createHash('sha256').update(value).digest('hex')\n}\n\nfunction toSelect(fields: string[]) {\n const select: Record = { id: true }\n for (const field of fields) select[field] = true\n return select\n}\n\nasync function main() {\n const runId = nowStamp()\n const backupDir = path.join(BACKUP_ROOT, runId)\n await fs.mkdir(backupDir, { recursive: true })\n\n const allRows: Array<{\n runId: string\n tableName: string\n rowId: string\n fieldName: string\n legacyValue: string\n checksum: string\n }> = []\n\n for (const mapping of MEDIA_MODEL_MAPPINGS) {\n const model = prismaDynamic[mapping.model]\n if (!model) continue\n\n const select = toSelect(mapping.fields.map((f) => f.legacyField))\n let cursor: string | null = null\n\n while (true) {\n const page = await model.findMany({\n select,\n ...(cursor\n ? {\n cursor: { id: cursor },\n skip: 1,\n }\n : {}),\n orderBy: { id: 'asc' },\n take: BATCH_SIZE,\n })\n if (!page.length) break\n\n for (const row of page) {\n for (const field of mapping.fields) {\n const value = row[field.legacyField]\n if (typeof value !== 'string' || !value.trim()) continue\n allRows.push({\n runId,\n tableName: mapping.tableName,\n rowId: String(row.id),\n fieldName: field.legacyField,\n legacyValue: value,\n checksum: checksum(value),\n })\n }\n }\n\n cursor = String(page[page.length - 1].id)\n }\n }\n\n if (allRows.length > 0) {\n try {\n const backupModel = prismaDynamic.legacyMediaRefBackup\n if (!backupModel?.createMany) {\n throw new Error('Prisma model not found: legacyMediaRefBackup')\n }\n for (let i = 0; i < allRows.length; i += 1000) {\n const chunk = allRows.slice(i, i + 1000)\n await backupModel.createMany({ data: chunk })\n }\n } catch (error) {\n const message = error instanceof Error ? error.message : String(error)\n _ulogError('[media-archive-legacy-refs] db backup table unavailable, fallback to file snapshot only', message)\n }\n }\n\n const snapshotPath = path.join(backupDir, 'legacy-media-refs.json')\n await fs.writeFile(snapshotPath, JSON.stringify(allRows, null, 2), 'utf8')\n const snapshotHash = checksum(await fs.readFile(snapshotPath, 'utf8'))\n\n const summary = {\n runId,\n createdAt: new Date().toISOString(),\n backupDir,\n archivedCount: allRows.length,\n snapshotFile: path.basename(snapshotPath),\n snapshotSha256: snapshotHash,\n }\n\n await fs.writeFile(path.join(backupDir, 'legacy-media-refs-summary.json'), JSON.stringify(summary, null, 2), 'utf8')\n\n _ulogInfo(`[media-archive-legacy-refs] runId=${runId}`)\n _ulogInfo(`[media-archive-legacy-refs] archived=${allRows.length}`)\n _ulogInfo(`[media-archive-legacy-refs] snapshot=${snapshotPath}`)\n}\n\nmain()\n .catch((error) => {\n _ulogError('[media-archive-legacy-refs] failed:', error)\n process.exitCode = 1\n })\n .finally(async () => {\n await prisma.$disconnect()\n })\n"
},
{
"path": "scripts/media-backfill-refs.ts",
"content": "import { logInfo as _ulogInfo, logError as _ulogError } from '@/lib/logging/core'\nimport { prisma } from '@/lib/prisma'\nimport { resolveMediaRefFromLegacyValue } from '@/lib/media/service'\nimport { MEDIA_MODEL_MAPPINGS } from './media-mapping'\n\nconst BATCH_SIZE = 200\ntype DynamicModel = {\n findMany: (args: unknown) => Promise>>\n update: (args: unknown) => Promise\n}\nconst prismaDynamic = prisma as unknown as Record\n\nfunction toSelect(fields: string[]) {\n const select: Record = { id: true }\n for (const field of fields) select[field] = true\n return select\n}\n\nasync function backfillModel(mapping: (typeof MEDIA_MODEL_MAPPINGS)[number]) {\n const model = prismaDynamic[mapping.model]\n if (!model) {\n throw new Error(`Prisma model not found: ${mapping.model}`)\n }\n\n const selectFields = mapping.fields.flatMap((f) => [f.legacyField, f.mediaIdField])\n const select = toSelect(selectFields)\n\n let cursor: string | null = null\n let scanned = 0\n let updated = 0\n\n try {\n while (true) {\n const rows = await model.findMany({\n select,\n ...(cursor\n ? {\n cursor: { id: cursor },\n skip: 1,\n }\n : {}),\n orderBy: { id: 'asc' },\n take: BATCH_SIZE,\n })\n\n if (!rows.length) break\n\n for (const row of rows) {\n scanned += 1\n const patch: Record = {}\n\n for (const field of mapping.fields) {\n const mediaId = row[field.mediaIdField]\n const legacyValue = row[field.legacyField]\n if (mediaId || typeof legacyValue !== 'string' || !legacyValue.trim()) {\n continue\n }\n\n const media = await resolveMediaRefFromLegacyValue(legacyValue)\n if (!media) continue\n patch[field.mediaIdField] = media.id\n }\n\n if (Object.keys(patch).length > 0) {\n await model.update({\n where: { id: String(row.id) },\n data: patch,\n })\n updated += 1\n }\n }\n\n cursor = String(rows[rows.length - 1].id)\n }\n } catch (error) {\n const message = error instanceof Error ? error.message : String(error)\n if (message.includes('does not exist') || message.includes('Unknown column')) {\n _ulogError(\n `[media-backfill-refs] skip ${mapping.tableName}: migration columns not available yet`,\n message,\n )\n return { scanned: 0, updated: 0, skipped: true }\n }\n throw error\n }\n\n return { scanned, updated, skipped: false }\n}\n\nasync function main() {\n const startedAt = new Date()\n _ulogInfo(`[media-backfill-refs] started at ${startedAt.toISOString()}`)\n\n let totalScanned = 0\n let totalUpdated = 0\n\n for (const mapping of MEDIA_MODEL_MAPPINGS) {\n const result = await backfillModel(mapping)\n totalScanned += result.scanned\n totalUpdated += result.updated\n if (result.skipped) {\n _ulogInfo(`[media-backfill-refs] ${mapping.tableName}: skipped (run add-only DB migration first)`)\n } else {\n _ulogInfo(\n `[media-backfill-refs] ${mapping.tableName}: scanned=${result.scanned} updatedRows=${result.updated}`,\n )\n }\n }\n\n _ulogInfo(\n `[media-backfill-refs] done scanned=${totalScanned} updatedRows=${totalUpdated} durationMs=${Date.now() - startedAt.getTime()}`,\n )\n}\n\nmain()\n .catch((error) => {\n _ulogError('[media-backfill-refs] failed:', error)\n process.exitCode = 1\n })\n .finally(async () => {\n await prisma.$disconnect()\n })\n"
},
{
"path": "scripts/media-build-unreferenced-index.ts",
"content": "import { logInfo as _ulogInfo, logError as _ulogError } from '@/lib/logging/core'\nimport { promises as fs } from 'node:fs'\nimport path from 'node:path'\nimport COS from 'cos-nodejs-sdk-v5'\nimport { prisma } from '@/lib/prisma'\nimport { resolveStorageKeyFromMediaValue } from '@/lib/media/service'\nimport { MEDIA_MODEL_MAPPINGS } from './media-mapping'\n\ntype StorageEntry = {\n key: string\n sizeBytes: number\n lastModified: string | null\n}\ntype CosBucketPage = {\n Contents?: Array<{ Key: string; Size?: string | number; LastModified?: string }>\n IsTruncated?: string | boolean\n NextMarker?: string\n}\ntype DynamicModel = {\n findMany: (args: unknown) => Promise>>\n}\nconst prismaDynamic = prisma as unknown as Record\n\nconst BACKUP_ROOT = path.join(process.cwd(), 'data', 'migration-backups')\n\nfunction nowStamp() {\n return new Date().toISOString().replace(/[:.]/g, '-')\n}\n\nasync function listLocalObjects(): Promise {\n const uploadDir = process.env.UPLOAD_DIR || './data/uploads'\n const rootDir = path.isAbsolute(uploadDir) ? uploadDir : path.join(process.cwd(), uploadDir)\n const exists = await fs.stat(rootDir).then(() => true).catch(() => false)\n if (!exists) return []\n\n const rows: StorageEntry[] = []\n const queue = ['']\n\n while (queue.length > 0) {\n const rel = queue.shift() as string\n const full = path.join(rootDir, rel)\n const entries = await fs.readdir(full, { withFileTypes: true })\n for (const entry of entries) {\n const childRel = path.join(rel, entry.name)\n if (entry.isDirectory()) {\n queue.push(childRel)\n continue\n }\n if (!entry.isFile()) continue\n const stat = await fs.stat(path.join(rootDir, childRel))\n rows.push({\n key: childRel.split(path.sep).join('/'),\n sizeBytes: stat.size,\n lastModified: stat.mtime.toISOString(),\n })\n }\n }\n\n return rows\n}\n\nasync function listCosObjects(): Promise {\n const secretId = process.env.COS_SECRET_ID\n const secretKey = process.env.COS_SECRET_KEY\n const bucket = process.env.COS_BUCKET\n const region = process.env.COS_REGION\n\n if (!secretId || !secretKey || !bucket || !region) {\n throw new Error('Missing COS env: COS_SECRET_ID/COS_SECRET_KEY/COS_BUCKET/COS_REGION')\n }\n\n const cos = new COS({ SecretId: secretId, SecretKey: secretKey, Timeout: 60_000 })\n const rows: StorageEntry[] = []\n let marker = ''\n\n while (true) {\n const page = await new Promise((resolve, reject) => {\n cos.getBucket(\n {\n Bucket: bucket,\n Region: region,\n Marker: marker,\n MaxKeys: 1000,\n },\n (err, data) => (err ? reject(err) : resolve(data as unknown as CosBucketPage)),\n )\n })\n\n const contents = page.Contents || []\n for (const item of contents) {\n rows.push({\n key: item.Key,\n sizeBytes: Number(item.Size || 0),\n lastModified: item.LastModified || null,\n })\n }\n\n const truncated = String(page.IsTruncated || 'false') === 'true'\n if (!truncated) break\n const nextMarker = typeof page.NextMarker === 'string' ? page.NextMarker : ''\n marker = nextMarker || (contents.length ? contents[contents.length - 1].Key : '')\n if (!marker) break\n }\n\n return rows\n}\n\nasync function listStorageObjects() {\n const storageType = process.env.STORAGE_TYPE || 'cos'\n if (storageType === 'local') {\n return { storageType, rows: await listLocalObjects() }\n }\n return { storageType, rows: await listCosObjects() }\n}\n\nasync function buildReferencedKeySet() {\n const refs = new Set()\n\n try {\n const mediaRows = await prismaDynamic.mediaObject.findMany({\n select: { storageKey: true },\n })\n for (const row of mediaRows) {\n if (typeof row.storageKey === 'string' && row.storageKey.trim()) refs.add(row.storageKey)\n }\n } catch (error) {\n const message = error instanceof Error ? error.message : String(error)\n _ulogError('[media-build-unreferenced-index] media_objects unavailable, fallback to legacy field scan', message)\n }\n\n for (const mapping of MEDIA_MODEL_MAPPINGS) {\n const model = prismaDynamic[mapping.model]\n if (!model) continue\n\n const select: Record = { id: true }\n for (const field of mapping.fields) select[field.legacyField] = true\n\n let cursor: string | null = null\n while (true) {\n const rows = await model.findMany({\n select,\n ...(cursor\n ? {\n cursor: { id: cursor },\n skip: 1,\n }\n : {}),\n orderBy: { id: 'asc' },\n take: 500,\n })\n if (!rows.length) break\n\n for (const row of rows) {\n for (const field of mapping.fields) {\n const value = row[field.legacyField]\n if (typeof value !== 'string' || !value.trim()) continue\n const key = await resolveStorageKeyFromMediaValue(value)\n if (key) refs.add(key)\n }\n }\n\n cursor = String(rows[rows.length - 1].id)\n }\n }\n\n return refs\n}\n\nasync function main() {\n const stamp = nowStamp()\n const backupDir = path.join(BACKUP_ROOT, stamp)\n await fs.mkdir(backupDir, { recursive: true })\n\n const referenced = await buildReferencedKeySet()\n const storage = await listStorageObjects()\n const unreferenced = storage.rows.filter((row) => !referenced.has(row.key))\n\n const output = {\n createdAt: new Date().toISOString(),\n storageType: storage.storageType,\n totalStorageObjects: storage.rows.length,\n referencedKeyCount: referenced.size,\n unreferencedCount: unreferenced.length,\n objects: unreferenced,\n }\n\n const filePath = path.join(backupDir, 'unreferenced-storage-objects-index.json')\n await fs.writeFile(filePath, JSON.stringify(output, null, 2), 'utf8')\n\n _ulogInfo(`[media-build-unreferenced-index] storageType=${storage.storageType}`)\n _ulogInfo(`[media-build-unreferenced-index] total=${storage.rows.length} unreferenced=${unreferenced.length}`)\n _ulogInfo(`[media-build-unreferenced-index] output=${filePath}`)\n}\n\nmain()\n .catch((error) => {\n _ulogError('[media-build-unreferenced-index] failed:', error)\n process.exitCode = 1\n })\n .finally(async () => {\n await prisma.$disconnect()\n })\n"
},
{
"path": "scripts/media-mapping.ts",
"content": "export type MediaFieldMapping = {\n legacyField: string\n mediaIdField: string\n}\n\nexport type MediaModelMapping = {\n model: string\n tableName: string\n fields: MediaFieldMapping[]\n}\n\nexport const MEDIA_MODEL_MAPPINGS: MediaModelMapping[] = [\n {\n model: 'characterAppearance',\n tableName: 'character_appearances',\n fields: [{ legacyField: 'imageUrl', mediaIdField: 'imageMediaId' }],\n },\n {\n model: 'locationImage',\n tableName: 'location_images',\n fields: [{ legacyField: 'imageUrl', mediaIdField: 'imageMediaId' }],\n },\n {\n model: 'novelPromotionCharacter',\n tableName: 'novel_promotion_characters',\n fields: [{ legacyField: 'customVoiceUrl', mediaIdField: 'customVoiceMediaId' }],\n },\n {\n model: 'novelPromotionEpisode',\n tableName: 'novel_promotion_episodes',\n fields: [{ legacyField: 'audioUrl', mediaIdField: 'audioMediaId' }],\n },\n {\n model: 'novelPromotionPanel',\n tableName: 'novel_promotion_panels',\n fields: [\n { legacyField: 'imageUrl', mediaIdField: 'imageMediaId' },\n { legacyField: 'videoUrl', mediaIdField: 'videoMediaId' },\n { legacyField: 'lipSyncVideoUrl', mediaIdField: 'lipSyncVideoMediaId' },\n { legacyField: 'sketchImageUrl', mediaIdField: 'sketchImageMediaId' },\n { legacyField: 'previousImageUrl', mediaIdField: 'previousImageMediaId' },\n ],\n },\n {\n model: 'novelPromotionShot',\n tableName: 'novel_promotion_shots',\n fields: [{ legacyField: 'imageUrl', mediaIdField: 'imageMediaId' }],\n },\n {\n model: 'supplementaryPanel',\n tableName: 'supplementary_panels',\n fields: [{ legacyField: 'imageUrl', mediaIdField: 'imageMediaId' }],\n },\n {\n model: 'novelPromotionVoiceLine',\n tableName: 'novel_promotion_voice_lines',\n fields: [{ legacyField: 'audioUrl', mediaIdField: 'audioMediaId' }],\n },\n {\n model: 'voicePreset',\n tableName: 'voice_presets',\n fields: [{ legacyField: 'audioUrl', mediaIdField: 'audioMediaId' }],\n },\n {\n model: 'globalCharacter',\n tableName: 'global_characters',\n fields: [{ legacyField: 'customVoiceUrl', mediaIdField: 'customVoiceMediaId' }],\n },\n {\n model: 'globalCharacterAppearance',\n tableName: 'global_character_appearances',\n fields: [\n { legacyField: 'imageUrl', mediaIdField: 'imageMediaId' },\n { legacyField: 'previousImageUrl', mediaIdField: 'previousImageMediaId' },\n ],\n },\n {\n model: 'globalLocationImage',\n tableName: 'global_location_images',\n fields: [\n { legacyField: 'imageUrl', mediaIdField: 'imageMediaId' },\n { legacyField: 'previousImageUrl', mediaIdField: 'previousImageMediaId' },\n ],\n },\n {\n model: 'globalVoice',\n tableName: 'global_voices',\n fields: [{ legacyField: 'customVoiceUrl', mediaIdField: 'customVoiceMediaId' }],\n },\n]\n"
},
{
"path": "scripts/media-restore-dry-run.ts",
"content": "import { logInfo as _ulogInfo, logError as _ulogError } from '@/lib/logging/core'\nimport { promises as fs } from 'node:fs'\nimport path from 'node:path'\nimport { prisma } from '@/lib/prisma'\n\nconst BACKUP_ROOT = path.join(process.cwd(), 'data', 'migration-backups')\n\ntype CountMap = Record\n\nasync function findLatestBackupDir() {\n const exists = await fs.stat(BACKUP_ROOT).then(() => true).catch(() => false)\n if (!exists) {\n throw new Error(`Backup root not found: ${BACKUP_ROOT}`)\n }\n const dirs = (await fs.readdir(BACKUP_ROOT, { withFileTypes: true }))\n .filter((d) => d.isDirectory())\n .map((d) => d.name)\n .sort()\n const validDirs: string[] = []\n for (const dir of dirs) {\n const metadataPath = path.join(BACKUP_ROOT, dir, 'metadata.json')\n const exists = await fs.stat(metadataPath).then(() => true).catch(() => false)\n if (exists) validDirs.push(dir)\n }\n\n if (!validDirs.length) {\n throw new Error(`No backup directories found in ${BACKUP_ROOT}`)\n }\n return path.join(BACKUP_ROOT, validDirs[validDirs.length - 1])\n}\n\nasync function readExpectedCounts(backupDir: string): Promise {\n const metadataPath = path.join(backupDir, 'metadata.json')\n const raw = await fs.readFile(metadataPath, 'utf8')\n const parsed = JSON.parse(raw)\n return (parsed.tableCounts || {}) as CountMap\n}\n\nasync function currentCounts(): Promise {\n const entries: Array<[string, string]> = [\n ['projects', 'projects'],\n ['novel_promotion_projects', 'novel_promotion_projects'],\n ['novel_promotion_episodes', 'novel_promotion_episodes'],\n ['novel_promotion_panels', 'novel_promotion_panels'],\n ['novel_promotion_voice_lines', 'novel_promotion_voice_lines'],\n ['global_characters', 'global_characters'],\n ['global_character_appearances', 'global_character_appearances'],\n ['global_locations', 'global_locations'],\n ['global_location_images', 'global_location_images'],\n ['global_voices', 'global_voices'],\n ['tasks', 'tasks'],\n ['task_events', 'task_events'],\n ]\n\n const resolved = await Promise.all(entries.map(async ([name, tableName]) => {\n const rows = (await prisma.$queryRawUnsafe(\n `SELECT COUNT(*) AS c FROM \\`${tableName}\\``,\n )) as Array>\n const raw = rows[0] || {}\n const firstValue = Object.values(raw)[0]\n const count = Number(firstValue || 0)\n return [name, Number.isFinite(count) ? count : 0] as const\n }))\n const out: CountMap = {}\n for (const [name, count] of resolved) out[name] = count\n return out\n}\n\nfunction printDiff(expected: CountMap, actual: CountMap) {\n const keys = [...new Set([...Object.keys(expected), ...Object.keys(actual)])].sort()\n let hasDiff = false\n\n _ulogInfo('table\\texpected\\tactual\\tdelta')\n for (const key of keys) {\n const e = expected[key] ?? 0\n const a = actual[key] ?? 0\n const d = a - e\n if (d !== 0) hasDiff = true\n _ulogInfo(`${key}\\t${e}\\t${a}\\t${d >= 0 ? '+' : ''}${d}`)\n }\n\n return hasDiff\n}\n\nasync function main() {\n const explicit = process.argv.find((arg) => arg.startsWith('--backup='))\n const backupDir = explicit ? path.resolve(explicit.split('=')[1]) : await findLatestBackupDir()\n\n _ulogInfo(`[media-restore-dry-run] backupDir=${backupDir}`)\n\n const expected = await readExpectedCounts(backupDir)\n const actual = await currentCounts()\n const hasDiff = printDiff(expected, actual)\n\n if (hasDiff) {\n _ulogInfo('[media-restore-dry-run] drift detected (dry-run only, no writes executed).')\n process.exitCode = 2\n return\n }\n\n _ulogInfo('[media-restore-dry-run] ok: counts match expected snapshot.')\n}\n\nmain()\n .catch((error) => {\n _ulogError('[media-restore-dry-run] failed:', error)\n process.exitCode = 1\n })\n .finally(async () => {\n await prisma.$disconnect()\n })\n"
},
{
"path": "scripts/media-safety-backup.ts",
"content": "import { logInfo as _ulogInfo, logError as _ulogError } from '@/lib/logging/core'\nimport { createHash } from 'node:crypto'\nimport { promises as fs } from 'node:fs'\nimport path from 'node:path'\nimport COS from 'cos-nodejs-sdk-v5'\nimport { prisma } from '@/lib/prisma'\n\ntype SnapshotTask = {\n name: string\n tableName: string\n}\n\ntype StorageIndexRow = {\n key: string\n hash: string | null\n sizeBytes: number\n lastModified: string | null\n}\n\ntype CosBucketPage = {\n Contents?: Array<{\n Key: string\n ETag?: string\n Size?: string | number\n LastModified?: string\n }>\n IsTruncated?: string | boolean\n NextMarker?: string\n}\n\nconst BACKUP_ROOT = path.join(process.cwd(), 'data', 'migration-backups')\n\nfunction nowStamp() {\n return new Date().toISOString().replace(/[:.]/g, '-')\n}\n\nfunction toJson(value: unknown) {\n return JSON.stringify(\n value,\n (_key, val) => (typeof val === 'bigint' ? String(val) : val),\n 2,\n )\n}\n\nasync function writeJson(filePath: string, data: unknown) {\n await fs.writeFile(filePath, toJson(data), 'utf8')\n}\n\nfunction sha256Text(input: string) {\n return createHash('sha256').update(input).digest('hex')\n}\n\nfunction resolveDatabaseFilePath(databaseUrl: string | undefined): string | null {\n if (!databaseUrl) return null\n if (databaseUrl.startsWith('file:')) {\n const raw = databaseUrl.slice('file:'.length)\n if (!raw) return null\n return path.isAbsolute(raw) ? raw : path.join(process.cwd(), raw)\n }\n return null\n}\n\nasync function listLocalFilesRecursively(rootDir: string, prefix = ''): Promise {\n const fullDir = path.join(rootDir, prefix)\n const entries = await fs.readdir(fullDir, { withFileTypes: true })\n const out: StorageIndexRow[] = []\n\n for (const entry of entries) {\n const rel = path.join(prefix, entry.name)\n if (entry.isDirectory()) {\n out.push(...(await listLocalFilesRecursively(rootDir, rel)))\n continue\n }\n if (!entry.isFile()) continue\n\n const filePath = path.join(rootDir, rel)\n const stat = await fs.stat(filePath)\n const buf = await fs.readFile(filePath)\n out.push({\n key: rel.split(path.sep).join('/'),\n hash: createHash('sha256').update(buf).digest('hex'),\n sizeBytes: stat.size,\n lastModified: stat.mtime.toISOString(),\n })\n }\n\n return out\n}\n\nasync function listCosObjects(): Promise {\n const secretId = process.env.COS_SECRET_ID\n const secretKey = process.env.COS_SECRET_KEY\n const bucket = process.env.COS_BUCKET\n const region = process.env.COS_REGION\n\n if (!secretId || !secretKey || !bucket || !region) {\n throw new Error('Missing COS env: COS_SECRET_ID/COS_SECRET_KEY/COS_BUCKET/COS_REGION')\n }\n\n const cos = new COS({ SecretId: secretId, SecretKey: secretKey, Timeout: 60_000 })\n const out: StorageIndexRow[] = []\n let marker = ''\n\n while (true) {\n const page = await new Promise((resolve, reject) => {\n cos.getBucket(\n {\n Bucket: bucket,\n Region: region,\n Marker: marker,\n MaxKeys: 1000,\n },\n (err, data) => (err ? reject(err) : resolve((data || {}) as CosBucketPage)),\n )\n })\n\n const contents = page.Contents || []\n for (const item of contents) {\n out.push({\n key: item.Key,\n hash: item.ETag ? String(item.ETag).replaceAll('\"', '') : null,\n sizeBytes: Number(item.Size || 0),\n lastModified: item.LastModified || null,\n })\n }\n\n const truncated = String(page.IsTruncated || 'false') === 'true'\n if (!truncated) break\n marker = page.NextMarker || (contents.length ? contents[contents.length - 1].Key : '')\n if (!marker) break\n }\n\n return out\n}\n\nasync function buildStorageIndex(): Promise<{ storageType: string; rows: StorageIndexRow[] }> {\n const storageType = process.env.STORAGE_TYPE || 'cos'\n if (storageType === 'local') {\n const uploadDir = process.env.UPLOAD_DIR || './data/uploads'\n const rootDir = path.isAbsolute(uploadDir) ? uploadDir : path.join(process.cwd(), uploadDir)\n const exists = await fs.stat(rootDir).then(() => true).catch(() => false)\n if (!exists) {\n return { storageType, rows: [] }\n }\n const rows = await listLocalFilesRecursively(rootDir)\n return { storageType, rows }\n }\n\n const rows = await listCosObjects()\n return { storageType, rows }\n}\n\nasync function snapshotTables(backupDir: string) {\n const tasks: SnapshotTask[] = [\n { name: 'projects', tableName: 'projects' },\n { name: 'novel_promotion_projects', tableName: 'novel_promotion_projects' },\n { name: 'novel_promotion_episodes', tableName: 'novel_promotion_episodes' },\n { name: 'novel_promotion_panels', tableName: 'novel_promotion_panels' },\n { name: 'novel_promotion_voice_lines', tableName: 'novel_promotion_voice_lines' },\n { name: 'global_characters', tableName: 'global_characters' },\n { name: 'global_character_appearances', tableName: 'global_character_appearances' },\n { name: 'global_locations', tableName: 'global_locations' },\n { name: 'global_location_images', tableName: 'global_location_images' },\n { name: 'global_voices', tableName: 'global_voices' },\n { name: 'tasks', tableName: 'tasks' },\n { name: 'task_events', tableName: 'task_events' },\n ]\n\n const counts: Record = {}\n for (const task of tasks) {\n const rows = (await prisma.$queryRawUnsafe(`SELECT * FROM \\`${task.tableName}\\``)) as unknown[]\n counts[task.name] = rows.length\n await writeJson(path.join(backupDir, `${task.name}.json`), rows)\n }\n\n return counts\n}\n\nasync function writeChecksums(backupDir: string) {\n const files = (await fs.readdir(backupDir)).sort()\n const sums: Record = {}\n\n for (const file of files) {\n const filePath = path.join(backupDir, file)\n const stat = await fs.stat(filePath)\n if (!stat.isFile()) continue\n const buf = await fs.readFile(filePath)\n sums[file] = createHash('sha256').update(buf).digest('hex')\n }\n\n await writeJson(path.join(backupDir, 'checksums.json'), sums)\n}\n\nasync function backupDbFile(backupDir: string) {\n const dbFile = resolveDatabaseFilePath(process.env.DATABASE_URL)\n if (!dbFile) return null\n\n const stat = await fs.stat(dbFile).catch(() => null)\n if (!stat || !stat.isFile()) return null\n\n const fileName = path.basename(dbFile)\n const target = path.join(backupDir, `db-file-${fileName}`)\n await fs.copyFile(dbFile, target)\n return path.basename(target)\n}\n\nasync function main() {\n const stamp = nowStamp()\n const backupDir = path.join(BACKUP_ROOT, stamp)\n await fs.mkdir(backupDir, { recursive: true })\n\n const meta: Record = {\n createdAt: new Date().toISOString(),\n backupDir,\n databaseUrl: process.env.DATABASE_URL || null,\n storageType: process.env.STORAGE_TYPE || 'cos',\n nodeEnv: process.env.NODE_ENV || null,\n }\n\n const copiedDbFile = await backupDbFile(backupDir)\n meta.copiedDbFile = copiedDbFile\n\n const tableCounts = await snapshotTables(backupDir)\n meta.tableCounts = tableCounts\n\n const storage = await buildStorageIndex()\n meta.storageType = storage.storageType\n meta.storageObjectCount = storage.rows.length\n await writeJson(path.join(backupDir, 'storage-object-index.json'), storage.rows)\n\n await writeChecksums(backupDir)\n meta.metadataChecksum = sha256Text(toJson(meta))\n await writeJson(path.join(backupDir, 'metadata.json'), meta)\n\n _ulogInfo(`[media-safety-backup] done: ${backupDir}`)\n _ulogInfo(`[media-safety-backup] tableCounts=${JSON.stringify(tableCounts)}`)\n _ulogInfo(`[media-safety-backup] storageObjects=${storage.rows.length}`)\n}\n\nmain()\n .catch((error) => {\n _ulogError('[media-safety-backup] failed:', error)\n process.exitCode = 1\n })\n .finally(async () => {\n await prisma.$disconnect()\n })\n"
},
{
"path": "scripts/migrate-cancelled-to-failed.ts",
"content": "import { prisma } from '@/lib/prisma'\n\nconst OLD_STATUS = 'cancelled'\nconst NEW_STATUS = 'failed'\nconst OLD_EVENT_TYPE = 'task.cancelled'\nconst NEW_EVENT_TYPE = 'task.failed'\nconst MIGRATION_ERROR_CODE = 'USER_CANCELLED'\nconst MIGRATION_ERROR_MESSAGE = '用户已停止任务。'\n\nfunction log(message: string) {\n process.stdout.write(`${message}\\n`)\n}\n\nfunction logError(message: string) {\n process.stderr.write(`${message}\\n`)\n}\n\nasync function main() {\n const totalTasks = await prisma.task.count({\n where: { status: OLD_STATUS },\n })\n const totalEvents = await prisma.taskEvent.count({\n where: { eventType: OLD_EVENT_TYPE },\n })\n\n log(`[migrate-cancelled-to-failed] matched tasks: ${totalTasks}`)\n log(`[migrate-cancelled-to-failed] matched events: ${totalEvents}`)\n if (totalTasks === 0 && totalEvents === 0) {\n log('[migrate-cancelled-to-failed] no rows to migrate')\n return\n }\n\n const taskEmptyMessageResult = await prisma.task.updateMany({\n where: {\n status: OLD_STATUS,\n OR: [{ errorMessage: null }, { errorMessage: '' }],\n },\n data: {\n status: NEW_STATUS,\n errorCode: MIGRATION_ERROR_CODE,\n errorMessage: MIGRATION_ERROR_MESSAGE,\n },\n })\n\n const taskResult = await prisma.task.updateMany({\n where: { status: OLD_STATUS },\n data: {\n status: NEW_STATUS,\n errorCode: MIGRATION_ERROR_CODE,\n },\n })\n\n const eventResult = await prisma.taskEvent.updateMany({\n where: { eventType: OLD_EVENT_TYPE },\n data: {\n eventType: NEW_EVENT_TYPE,\n },\n })\n\n log(`[migrate-cancelled-to-failed] updated tasks (empty message): ${taskEmptyMessageResult.count}`)\n log(`[migrate-cancelled-to-failed] updated tasks (remaining): ${taskResult.count}`)\n log(`[migrate-cancelled-to-failed] updated events: ${eventResult.count}`)\n}\n\nmain()\n .catch((error) => {\n logError(`[migrate-cancelled-to-failed] failed: ${error instanceof Error ? error.stack || error.message : String(error)}`)\n process.exitCode = 1\n })\n .finally(async () => {\n await prisma.$disconnect()\n })\n"
},
{
"path": "scripts/migrate-image-urls-contract.ts",
"content": "import { logInfo as _ulogInfo, logError as _ulogError } from '@/lib/logging/core'\nimport { prisma } from '@/lib/prisma'\nimport { encodeImageUrls } from '@/lib/contracts/image-urls-contract'\n\ntype AppearanceRow = {\n id: string\n imageUrls: string | null\n previousImageUrls: string | null\n}\n\ntype DynamicModel = {\n findMany: (args: unknown) => Promise\n update: (args: unknown) => Promise\n}\n\ntype FieldName = 'imageUrls' | 'previousImageUrls'\n\ntype NormalizeResult = {\n next: string\n changed: boolean\n reason: 'ok' | 'null' | 'invalid_json' | 'not_array' | 'filtered_non_string' | 'normalized_json'\n}\n\ntype ModelStats = {\n scanned: number\n updatedRows: number\n changedFields: number\n reasons: Record\n}\n\nconst BATCH_SIZE = 200\nconst APPLY = process.argv.includes('--apply')\n\nconst MODELS: Array<{ name: string; model: string }> = [\n { name: 'CharacterAppearance', model: 'characterAppearance' },\n { name: 'GlobalCharacterAppearance', model: 'globalCharacterAppearance' },\n]\n\nconst prismaDynamic = prisma as unknown as Record\n\nfunction print(message: string) {\n process.stdout.write(`${message}\\n`)\n}\n\nfunction normalizeField(raw: string | null): NormalizeResult {\n if (raw === null) {\n return {\n next: encodeImageUrls([]),\n changed: true,\n reason: 'null',\n }\n }\n\n try {\n const parsed = JSON.parse(raw) as unknown\n if (!Array.isArray(parsed)) {\n return {\n next: encodeImageUrls([]),\n changed: true,\n reason: 'not_array',\n }\n }\n\n const stringOnly = parsed.filter((item): item is string => typeof item === 'string')\n const next = encodeImageUrls(stringOnly)\n\n if (parsed.length !== stringOnly.length) {\n return {\n next,\n changed: true,\n reason: 'filtered_non_string',\n }\n }\n\n if (raw !== next) {\n return {\n next,\n changed: true,\n reason: 'normalized_json',\n }\n }\n\n return {\n next,\n changed: false,\n reason: 'ok',\n }\n } catch {\n return {\n next: encodeImageUrls([]),\n changed: true,\n reason: 'invalid_json',\n }\n }\n}\n\nasync function migrateModel(modelName: string, modelKey: string) {\n const model = prismaDynamic[modelKey]\n if (!model) {\n throw new Error(`Prisma model not found: ${modelKey}`)\n }\n\n const stats: ModelStats = {\n scanned: 0,\n updatedRows: 0,\n changedFields: 0,\n reasons: {\n ok: 0,\n null: 0,\n invalid_json: 0,\n not_array: 0,\n filtered_non_string: 0,\n normalized_json: 0,\n },\n }\n\n const samples: Array<{ id: string; field: FieldName; reason: NormalizeResult['reason']; before: string | null; after: string }> = []\n\n let cursor: string | null = null\n\n while (true) {\n const rows = await model.findMany({\n select: {\n id: true,\n imageUrls: true,\n previousImageUrls: true,\n },\n ...(cursor\n ? {\n cursor: { id: cursor },\n skip: 1,\n }\n : {}),\n orderBy: { id: 'asc' },\n take: BATCH_SIZE,\n })\n\n if (rows.length === 0) break\n\n for (const row of rows) {\n stats.scanned += 1\n\n const imageUrlsResult = normalizeField(row.imageUrls)\n const previousImageUrlsResult = normalizeField(row.previousImageUrls)\n\n stats.reasons[imageUrlsResult.reason] += 1\n stats.reasons[previousImageUrlsResult.reason] += 1\n\n const data: Partial> = {}\n\n if (imageUrlsResult.changed) {\n data.imageUrls = imageUrlsResult.next\n stats.changedFields += 1\n if (samples.length < 20) {\n samples.push({\n id: row.id,\n field: 'imageUrls',\n reason: imageUrlsResult.reason,\n before: row.imageUrls,\n after: imageUrlsResult.next,\n })\n }\n }\n\n if (previousImageUrlsResult.changed) {\n data.previousImageUrls = previousImageUrlsResult.next\n stats.changedFields += 1\n if (samples.length < 20) {\n samples.push({\n id: row.id,\n field: 'previousImageUrls',\n reason: previousImageUrlsResult.reason,\n before: row.previousImageUrls,\n after: previousImageUrlsResult.next,\n })\n }\n }\n\n if (Object.keys(data).length > 0) {\n stats.updatedRows += 1\n if (APPLY) {\n await model.update({\n where: { id: row.id },\n data,\n })\n }\n }\n }\n\n cursor = rows[rows.length - 1]?.id || null\n }\n\n const summary = `[migrate-image-urls-contract] ${modelName}: scanned=${stats.scanned} updatedRows=${stats.updatedRows} changedFields=${stats.changedFields}`\n _ulogInfo(summary)\n print(summary)\n print(`[migrate-image-urls-contract] ${modelName}: reasons=${JSON.stringify(stats.reasons)}`)\n\n if (samples.length > 0) {\n print(`[migrate-image-urls-contract] ${modelName}: sampleChanges=${JSON.stringify(samples, null, 2)}`)\n }\n\n return stats\n}\n\nasync function main() {\n print(`[migrate-image-urls-contract] mode=${APPLY ? 'apply' : 'dry-run'}`)\n\n const totals = {\n scanned: 0,\n updatedRows: 0,\n changedFields: 0,\n }\n\n for (const target of MODELS) {\n const stats = await migrateModel(target.name, target.model)\n totals.scanned += stats.scanned\n totals.updatedRows += stats.updatedRows\n totals.changedFields += stats.changedFields\n }\n\n print(`[migrate-image-urls-contract] done scanned=${totals.scanned} updatedRows=${totals.updatedRows} changedFields=${totals.changedFields} mode=${APPLY ? 'apply' : 'dry-run'}`)\n}\n\nmain()\n .catch((error) => {\n _ulogError('[migrate-image-urls-contract] failed:', error)\n process.exitCode = 1\n })\n .finally(async () => {\n await prisma.$disconnect()\n })\n"
},
{
"path": "scripts/migrate-local-to-minio.ts",
"content": "#!/usr/bin/env npx tsx\n/**\n * 本地存储 → MinIO 迁移脚本\n * 使用 @aws-sdk/client-s3（项目已有依赖）\n * \n * 用法: npx tsx scripts/migrate-local-to-minio.ts\n */\n\nimport { S3Client, PutObjectCommand, HeadObjectCommand } from '@aws-sdk/client-s3'\nimport * as fs from 'fs/promises'\nimport * as path from 'path'\nimport { createReadStream } from 'fs'\n\n// ==================== 配置 ====================\nconst LOCAL_DIR = process.env.LOCAL_UPLOAD_DIR || './data/uploads'\nconst MINIO_ENDPOINT = process.env.MINIO_ENDPOINT || 'http://127.0.0.1:19000'\nconst MINIO_BUCKET = process.env.MINIO_BUCKET || 'waoowaoo'\nconst MINIO_REGION = process.env.MINIO_REGION || 'us-east-1'\nconst MINIO_ACCESS_KEY = process.env.MINIO_ACCESS_KEY || 'minioadmin'\nconst MINIO_SECRET_KEY = process.env.MINIO_SECRET_KEY || 'minioadmin'\nconst CONCURRENCY = parseInt(process.env.MIGRATE_CONCURRENCY || '10')\nconst DRY_RUN = process.env.MIGRATE_DRY_RUN === 'true'\n\n// ==================== S3 客户端 ====================\nconst s3 = new S3Client({\n endpoint: MINIO_ENDPOINT,\n region: MINIO_REGION,\n forcePathStyle: true,\n credentials: {\n accessKeyId: MINIO_ACCESS_KEY,\n secretAccessKey: MINIO_SECRET_KEY,\n },\n})\n\n// ==================== 工具函数 ====================\nfunction guessContentType(filename: string): string {\n const ext = path.extname(filename).toLowerCase()\n const types: Record = {\n '.jpg': 'image/jpeg',\n '.jpeg': 'image/jpeg',\n '.png': 'image/png',\n '.gif': 'image/gif',\n '.webp': 'image/webp',\n '.mp4': 'video/mp4',\n '.webm': 'video/webm',\n '.mp3': 'audio/mpeg',\n '.wav': 'audio/wav',\n '.ogg': 'audio/ogg',\n '.json': 'application/json',\n '.txt': 'text/plain',\n }\n return types[ext] || 'application/octet-stream'\n}\n\nfunction formatBytes(bytes: number): string {\n if (bytes === 0) return '0 B'\n const k = 1024\n const sizes = ['B', 'KB', 'MB', 'GB']\n const i = Math.floor(Math.log(bytes) / Math.log(k))\n return parseFloat((bytes / Math.pow(k, i)).toFixed(2)) + ' ' + sizes[i]\n}\n\n// ==================== 扫描本地文件 ====================\nasync function scanLocalFiles(dir: string, basePath = ''): Promise> {\n const files: Array<{ localPath: string; key: string; size: number }> = []\n\n try {\n const entries = await fs.readdir(dir, { withFileTypes: true })\n for (const entry of entries) {\n const fullPath = path.join(dir, entry.name)\n const relativePath = path.join(basePath, entry.name)\n\n if (entry.isDirectory()) {\n files.push(...await scanLocalFiles(fullPath, relativePath))\n } else {\n // 跳过隐藏文件\n if (entry.name.startsWith('.')) continue\n const stats = await fs.stat(fullPath)\n files.push({\n localPath: fullPath,\n key: relativePath.replace(/\\\\/g, '/'),\n size: stats.size,\n })\n }\n }\n } catch (err: unknown) {\n console.error(` ⚠️ 无法读取目录: ${dir}`, (err as Error).message)\n }\n\n return files\n}\n\n// ==================== 检查文件是否已存在 ====================\nasync function objectExists(key: string): Promise {\n try {\n await s3.send(new HeadObjectCommand({ Bucket: MINIO_BUCKET, Key: key }))\n return true\n } catch {\n return false\n }\n}\n\n// ==================== 上传文件 ====================\nasync function uploadFile(file: { localPath: string; key: string; size: number }): Promise<'success' | 'skipped' | 'error'> {\n // 检查是否已存在\n if (await objectExists(file.key)) {\n return 'skipped'\n }\n\n if (DRY_RUN) {\n console.log(` [DRY RUN] 将上传: ${file.key} (${formatBytes(file.size)})`)\n return 'skipped'\n }\n\n try {\n const body = await fs.readFile(file.localPath)\n await s3.send(new PutObjectCommand({\n Bucket: MINIO_BUCKET,\n Key: file.key,\n Body: body,\n ContentType: guessContentType(file.key),\n }))\n return 'success'\n } catch (err: unknown) {\n console.error(` ✗ 上传失败: ${file.key}`, (err as Error).message)\n return 'error'\n }\n}\n\n// ==================== 并行控制 ====================\nasync function runBatched(items: T[], concurrency: number, fn: (item: T) => Promise) {\n for (let i = 0; i < items.length; i += concurrency) {\n const batch = items.slice(i, i + concurrency)\n await Promise.all(batch.map(fn))\n }\n}\n\n// ==================== 主流程 ====================\nasync function main() {\n console.log()\n console.log('╔══════════════════════════════════════════════════════╗')\n console.log('║ Local Storage → MinIO Migration Tool ║')\n console.log('╚══════════════════════════════════════════════════════╝')\n console.log()\n console.log(` 📂 源目录: ${path.resolve(LOCAL_DIR)}`)\n console.log(` 🪣 目标桶: ${MINIO_ENDPOINT}/${MINIO_BUCKET}`)\n console.log(` ⚡ 并发数: ${CONCURRENCY}`)\n console.log(` 🔍 干运行: ${DRY_RUN}`)\n console.log()\n\n // 1. 扫描文件\n console.log('📦 扫描本地文件...')\n const files = await scanLocalFiles(LOCAL_DIR)\n\n if (files.length === 0) {\n console.log(' 没有需要迁移的文件')\n return\n }\n\n const totalSize = files.reduce((sum, f) => sum + f.size, 0)\n console.log(` 找到 ${files.length} 个文件, 总大小: ${formatBytes(totalSize)}`)\n console.log()\n\n // 2. 开始上传\n console.log('🚀 开始迁移...')\n const startTime = Date.now()\n let success = 0\n let skipped = 0\n let failed = 0\n let processed = 0\n\n await runBatched(files, CONCURRENCY, async (file) => {\n const result = await uploadFile(file)\n processed++\n\n if (result === 'success') {\n success++\n if (success % 50 === 0 || success <= 5) {\n console.log(` ✓ [${processed}/${files.length}] ${file.key} (${formatBytes(file.size)})`)\n }\n } else if (result === 'skipped') {\n skipped++\n } else {\n failed++\n }\n\n if (processed % 100 === 0) {\n const pct = ((processed / files.length) * 100).toFixed(1)\n console.log(` 📊 进度: ${pct}% (${processed}/${files.length}) | ✓${success} ⏭${skipped} ✗${failed}`)\n }\n })\n\n // 3. 结果\n const duration = ((Date.now() - startTime) / 1000).toFixed(1)\n console.log()\n console.log('╔══════════════════════════════════════════════════════╗')\n console.log('║ 迁移完成 ║')\n console.log('╠══════════════════════════════════════════════════════╣')\n console.log(`║ 总文件: ${String(files.length).padEnd(40)} ║`)\n console.log(`║ 成功: ${String(success).padEnd(40)} ║`)\n console.log(`║ 跳过: ${String(skipped).padEnd(40)} ║`)\n console.log(`║ 失败: ${String(failed).padEnd(40)} ║`)\n console.log(`║ 耗时: ${String(duration + 's').padEnd(40)} ║`)\n console.log(`║ 大小: ${formatBytes(totalSize).padEnd(40)} ║`)\n console.log('╚══════════════════════════════════════════════════════╝')\n\n if (failed > 0) {\n console.log()\n console.log('⚠️ 有文件上传失败，请重新运行脚本（已上传的会自动跳过）')\n process.exit(1)\n }\n}\n\nmain().catch(err => {\n console.error('迁移失败:', err)\n process.exit(1)\n})\n"
},
{
"path": "scripts/migrate-to-minio.sh",
"content": "#!/bin/bash\n#\n# 存储迁移快捷脚本\n# 用法: ./scripts/migrate-to-minio.sh [选项]\n#\n\nset -e\n\nSCRIPT_DIR=\"$(cd \"$(dirname \"${BASH_SOURCE[0]}\")\" && pwd)\"\nPROJECT_DIR=\"$(dirname \"$SCRIPT_DIR\")\"\ncd \"$PROJECT_DIR\"\n\n# 颜色输出\nRED='\\033[0;31m'\nGREEN='\\033[0;32m'\nYELLOW='\\033[1;33m'\nNC='\\033[0m' # No Color\n\necho -e \"${GREEN}══════════════════════════════════════════════════════════${NC}\"\necho -e \"${GREEN} Local Storage → MinIO Migration Tool${NC}\"\necho -e \"${GREEN}══════════════════════════════════════════════════════════${NC}\"\necho\n\n# 检查 MinIO 是否运行\nif ! curl -sf http://127.0.0.1:19000/minio/health/live >/devdev/null 2>&1; then\n echo -e \"${YELLOW}⚠ MinIO 未检测到在 127.0.0.1:19000${NC}\"\n echo \" 请先启动 MinIO: docker compose up -d minio\"\n echo\n read -p \"是否尝试自动启动 MinIO? [Y/n] \" -n 1 -r\n echo\n if [[ ! $REPLY =~ ^[Nn]$ ]]; then\n docker compose up -d minio\n echo -e \"${GREEN}✓ MinIO 启动中，等待 5 秒...${NC}\"\n sleep 5\n else\n exit 1\n fi\nfi\n\necho -e \"${GREEN}✓ MinIO 服务正常${NC}\"\necho\n\n# 检查本地数据目录\nif [ ! -d \"./data/uploads\" ]; then\n echo -e \"${YELLOW}⚠ 本地数据目录 ./data/uploads 不存在${NC}\"\n echo \" 无需迁移\"\n exit 0\nfi\n\nFILE_COUNT=$(find ./data/uploads -type f 2>/dev/null | wc -l)\nif [ \"$FILE_COUNT\" -eq 0 ]; then\n echo -e \"${YELLOW}⚠ 本地数据目录为空${NC}\"\n echo \" 无需迁移\"\n exit 0\nfi\n\necho \"本地文件数: $FILE_COUNT\"\necho\n\n# 运行干运行模式预览\necho -e \"${YELLOW}▶ 干运行预览 (Dry Run)...${NC}\"\nMIGRATE_DRY_RUN=true npx tsx scripts/migrate-to-minio.ts\necho\n\n# 确认执行\nread -p \"是否开始实际迁移? [y/N] \" -n 1 -r\necho\nif [[ ! $REPLY =~ ^[Yy]$ ]]; then\n echo -e \"${YELLOW}已取消迁移${NC}\"\n exit 0\nfi\n\necho\necho -e \"${GREEN}▶ 开始迁移...${NC}\"\nnpx tsx scripts/migrate-to-minio.ts\n\nif [ $? -eq 0 ]; then\n echo\n echo -e \"${GREEN}══════════════════════════════════════════════════════════${NC}\"\n echo -e \"${GREEN} 迁移成功完成!${NC}\"\n echo -e \"${GREEN}══════════════════════════════════════════════════════════${NC}\"\n echo\n echo \"后续步骤:\"\n echo \" 1. 验证 MinIO 控制台: http://127.0.0.1:19001\"\n echo \" 账号: minioadmin / minioadmin\"\n echo \" 2. 更新 .env: STORAGE_TYPE=minio\"\n echo \" 3. 重启应用: docker compose restart app\"\n echo \" 4. 测试图片/视频访问\"\n echo \" 5. 确认无误后删除本地数据: rm -rf ./data/uploads\"\n echo\nelse\n echo\n echo -e \"${RED}══════════════════════════════════════════════════════════${NC}\"\n echo -e \"${RED} 迁移失败${NC}\"\n echo -e \"${RED}══════════════════════════════════════════════════════════${NC}\"\n echo\n echo \"可重新运行继续迁移:\"\n echo \" ./scripts/migrate-to-minio.sh\"\n exit 1\nfi\n"
},
{
"path": "scripts/migrate-to-minio.ts",
"content": "#!/usr/bin/env node\n/**\n * 存储迁移脚本: Local → MinIO\n * \n * 用途: 将本地文件存储的数据无缝迁移到 MinIO 对象存储\n * 特点:\n * - 断点续传（记录已迁移文件）\n * - 校验和验证\n * - 原子性操作（失败可回滚）\n * - 并行上传加速\n */\n\nimport { Client as MinioClient } from 'minio'\nimport * as fs from 'fs/promises'\nimport * as path from 'path'\nimport { createHash } from 'crypto'\nimport { createReadStream } from 'fs'\n\n// ==================== 配置 ====================\nconst CONFIG = {\n // 源: 本地存储\n local: {\n baseDir: process.env.LOCAL_UPLOAD_DIR || './data/uploads',\n },\n // 目标: MinIO\n minio: {\n endPoint: process.env.MINIO_ENDPOINT?.replace(/^https?:\\/\\//, '') || '127.0.0.1',\n port: parseInt(process.env.MINIO_PORT || '9000'),\n useSSL: process.env.MINIO_USE_SSL === 'true',\n accessKey: process.env.MINIO_ACCESS_KEY || 'minioadmin',\n secretKey: process.env.MINIO_SECRET_KEY || 'minioadmin',\n bucket: process.env.MINIO_BUCKET || 'waoowaoo',\n region: process.env.MINIO_REGION || 'us-east-1',\n forcePathStyle: process.env.MINIO_FORCE_PATH_STYLE !== 'false',\n },\n // 迁移选项\n options: {\n concurrency: parseInt(process.env.MIGRATE_CONCURRENCY || '5'),\n dryRun: process.env.MIGRATE_DRY_RUN === 'true',\n resume: process.env.MIGRATE_RESUME !== 'false',\n progressFile: process.env.MIGRATE_PROGRESS_FILE || './scripts/.migrate-progress.json',\n logLevel: process.env.MIGRATE_LOG_LEVEL || 'info', // debug, info, warn, error\n }\n}\n\n// ==================== 日志 ====================\nconst LOG_LEVELS = { debug: 0, info: 1, warn: 2, error: 3 }\nfunction log(level: string, message: string, ...args: unknown[]) {\n if (LOG_LEVELS[level as keyof typeof LOG_LEVELS] >= LOG_LEVELS[CONFIG.options.logLevel as keyof typeof LOG_LEVELS]) {\n const timestamp = new Date().toISOString()\n console[level === 'error' ? 'error' : 'log'](`[${timestamp}] [${level.toUpperCase()}] ${message}`, ...args)\n }\n}\n\n// ==================== MinIO 客户端 ====================\nconst minioClient = new MinioClient({\n endPoint: CONFIG.minio.endPoint,\n port: CONFIG.minio.port,\n useSSL: CONFIG.minio.useSSL,\n accessKey: CONFIG.minio.accessKey,\n secretKey: CONFIG.minio.secretKey,\n region: CONFIG.minio.region,\n})\n\n// ==================== 文件扫描 ====================\nasync function scanLocalFiles(dir: string, basePath = ''): Promise> {\n const files: Array<{localPath: string, key: string, size: number, mtime: Date}> = []\n \n try {\n const entries = await fs.readdir(dir, { withFileTypes: true })\n \n for (const entry of entries) {\n const fullPath = path.join(dir, entry.name)\n const relativePath = path.join(basePath, entry.name)\n \n if (entry.isDirectory()) {\n const subFiles = await scanLocalFiles(fullPath, relativePath)\n files.push(...subFiles)\n } else {\n const stats = await fs.stat(fullPath)\n files.push({\n localPath: fullPath,\n key: relativePath.replace(/\\\\/g, '/'), // 统一使用正斜杠\n size: stats.size,\n mtime: stats.mtime,\n })\n }\n }\n } catch (err: unknown) {\n log('warn', `无法读取目录: ${dir}`, (err as Error).message)\n }\n \n return files\n}\n\n// ==================== 校验和 ====================\nasync function calculateHash(filePath: string): Promise {\n return new Promise((resolve, reject) => {\n const hash = createHash('md5')\n const stream = createReadStream(filePath)\n \n stream.on('data', chunk => hash.update(chunk))\n stream.on('end', () => resolve(hash.digest('hex')))\n stream.on('error', reject)\n })\n}\n\n// ==================== 进度管理 ====================\nasync function loadProgress(): Promise> {\n try {\n if (!CONFIG.options.resume) {\n return new Set()\n }\n const data = await fs.readFile(CONFIG.options.progressFile, 'utf-8')\n const progress = JSON.parse(data)\n return new Set(progress.migrated || [])\n } catch {\n return new Set()\n }\n}\n\nasync function saveProgress(migratedKeys: Set) {\n const progress = {\n updatedAt: new Date().toISOString(),\n migrated: Array.from(migratedKeys),\n }\n await fs.writeFile(CONFIG.options.progressFile, JSON.stringify(progress, null, 2))\n}\n\n// ==================== 存储桶检查/创建 ====================\nasync function ensureBucket() {\n log('info', `检查存储桶: ${CONFIG.minio.bucket}`)\n \n const exists = await minioClient.bucketExists(CONFIG.minio.bucket)\n if (!exists) {\n log('info', `创建存储桶: ${CONFIG.minio.bucket}`)\n await minioClient.makeBucket(CONFIG.minio.bucket, CONFIG.minio.region)\n \n // 设置存储桶为 public read (可选，根据需求)\n const policy = {\n Version: '2012-10-17',\n Statement: [\n {\n Effect: 'Allow',\n Principal: { AWS: ['*'] },\n Action: ['s3:GetObject'],\n Resource: [`arn:aws:s3:::${CONFIG.minio.bucket}/*`]\n }\n ]\n }\n await minioClient.setBucketPolicy(CONFIG.minio.bucket, JSON.stringify(policy))\n log('info', '存储桶访问策略已设置为公开读取')\n }\n}\n\n// ==================== 文件上传 ====================\nasync function uploadFile(fileInfo: {localPath: string, key: string, size: number}, migratedKeys: Set): Promise<{status: string, key: string, size?: number, error?: string}> {\n const { localPath, key, size } = fileInfo\n \n // 检查是否已迁移\n if (migratedKeys.has(key)) {\n log('debug', `跳过已迁移: ${key}`)\n return { status: 'skipped', key }\n }\n \n if (CONFIG.options.dryRun) {\n log('info', `[DRY RUN] 将上传: ${key} (${formatBytes(size)})`)\n return { status: 'dry_run', key }\n }\n \n try {\n // 计算本地文件 MD5\n const localHash = await calculateHash(localPath)\n \n // 上传文件\n const fileStream = createReadStream(localPath)\n await minioClient.putObject(CONFIG.minio.bucket, key, fileStream, size, {\n 'Content-Type': guessContentType(key),\n 'X-Amz-Meta-Original-Hash': localHash,\n })\n \n // 验证上传\n const stat = await minioClient.statObject(CONFIG.minio.bucket, key)\n \n // 记录迁移成功\n migratedKeys.add(key)\n \n log('info', `✓ 上传成功: ${key} (${formatBytes(size)})`)\n return { status: 'success', key, size }\n \n } catch (err: unknown) {\n log('error', `✗ 上传失败: ${key}`, (err as Error).message)\n return { status: 'error', key, error: (err as Error).message }\n }\n}\n\n// ==================== 内容类型猜测 ====================\nfunction guessContentType(filename: string): string {\n const ext = path.extname(filename).toLowerCase()\n const types: Record = {\n '.jpg': 'image/jpeg',\n '.jpeg': 'image/jpeg',\n '.png': 'image/png',\n '.gif': 'image/gif',\n '.webp': 'image/webp',\n '.mp4': 'video/mp4',\n '.webm': 'video/webm',\n '.mp3': 'audio/mpeg',\n '.wav': 'audio/wav',\n '.json': 'application/json',\n '.txt': 'text/plain',\n }\n return types[ext] || 'application/octet-stream'\n}\n\n// ==================== 字节格式化 ====================\nfunction formatBytes(bytes: number): string {\n if (bytes === 0) return '0 B'\n const k = 1024\n const sizes = ['B', 'KB', 'MB', 'GB', 'TB']\n const i = Math.floor(Math.log(bytes) / Math.log(k))\n return parseFloat((bytes / Math.pow(k, i)).toFixed(2)) + ' ' + sizes[i]\n}\n\n// ==================== 并行任务控制 ====================\nasync function runWithConcurrency(tasks: Array<() => Promise>, concurrency: number): Promise {\n const results: T[] = []\n const executing: Promise[] = []\n \n for (const task of tasks) {\n const promise = task().then(result => {\n results.push(result)\n })\n executing.push(promise)\n \n if (executing.length >= concurrency) {\n await Promise.race(executing)\n executing.splice(executing.findIndex(p => p === promise), 1)\n }\n }\n \n await Promise.all(executing)\n return results\n}\n\n// ==================== 主流程 ====================\nasync function main() {\n console.log('╔══════════════════════════════════════════════════════════╗')\n console.log('║ Local Storage → MinIO Migration Tool ║')\n console.log('╚══════════════════════════════════════════════════════════╝')\n console.log()\n \n log('info', '配置信息:')\n log('info', ` 本地目录: ${path.resolve(CONFIG.local.baseDir)}`)\n log('info', ` MinIO: ${CONFIG.minio.endPoint}:${CONFIG.minio.port}/${CONFIG.minio.bucket}`)\n log('info', ` 并发数: ${CONFIG.options.concurrency}`)\n log('info', ` 干运行: ${CONFIG.options.dryRun}`)\n log('info', ` 断点续传: ${CONFIG.options.resume}`)\n console.log()\n \n // 1. 扫描本地文件\n log('info', '扫描本地文件...')\n const files = await scanLocalFiles(CONFIG.local.baseDir)\n log('info', `找到 ${files.length} 个文件`)\n \n if (files.length === 0) {\n log('info', '没有需要迁移的文件')\n return\n }\n \n const totalSize = files.reduce((sum, f) => sum + f.size, 0)\n log('info', `总大小: ${formatBytes(totalSize)}`)\n console.log()\n \n // 2. 加载进度\n const migratedKeys = await loadProgress()\n log('info', `已迁移: ${migratedKeys.size} 个文件`)\n \n // 3. 确保存储桶存在\n await ensureBucket()\n \n // 4. 执行迁移\n const startTime = Date.now()\n let processed = 0\n let success = 0\n let failed = 0\n let skipped = 0\n \n const uploadTasks = files.map(file => async () => {\n const result = await uploadFile(file, migratedKeys)\n processed++\n \n if (result.status === 'success') success++\n else if (result.status === 'error') failed++\n else if (result.status === 'skipped') skipped++\n \n // 每 10 个文件保存一次进度\n if (processed % 10 === 0) {\n await saveProgress(migratedKeys)\n const progress = ((processed / files.length) * 100).toFixed(1)\n log('info', `进度: ${progress}% (${processed}/${files.length})`)\n }\n \n return result\n })\n \n await runWithConcurrency(uploadTasks, CONFIG.options.concurrency)\n \n // 5. 保存最终进度\n await saveProgress(migratedKeys)\n \n // 6. 报告\n const duration = ((Date.now() - startTime) / 1000).toFixed(1)\n console.log()\n console.log('╔══════════════════════════════════════════════════════════╗')\n console.log('║ 迁移完成 ║')\n console.log('╠══════════════════════════════════════════════════════════╣')\n console.log(`║ 总文件数: ${String(files.length).padEnd(39)} ║`)\n console.log(`║ 成功: ${String(success).padEnd(39)} ║`)\n console.log(`║ 失败: ${String(failed).padEnd(39)} ║`)\n console.log(`║ 跳过: ${String(skipped).padEnd(39)} ║`)\n console.log(`║ 耗时: ${String(duration + 's').padEnd(39)} ║`)\n console.log('╚══════════════════════════════════════════════════════════╝')\n \n // 7. 后续步骤提示\n console.log()\n console.log('📋 后续步骤:')\n console.log(' 1. 验证 MinIO 中的文件: mc ls local/waoowaoo')\n console.log(' 2. 更新 .env: STORAGE_TYPE=minio')\n console.log(' 3. 重启应用: docker compose restart app')\n console.log(' 4. 测试图片/视频访问是否正常')\n console.log(' 5. 确认无误后可删除本地文件: rm -rf ./data/uploads')\n \n if (failed > 0) {\n process.exit(1)\n }\n}\n\n// 运行\nmain().catch(err => {\n log('error', '迁移失败:', err)\n process.exit(1)\n})\n"
},
{
"path": "scripts/migrations/migrate-capability-selections.ts",
"content": "import { prisma } from '@/lib/prisma'\nimport {\n parseModelKeyStrict,\n type CapabilitySelections,\n type CapabilityValue,\n} from '@/lib/model-config-contract'\nimport { findBuiltinCapabilities } from '@/lib/model-capabilities/catalog'\n\nconst APPLY = process.argv.includes('--apply')\n\nconst USER_IMAGE_MODEL_FIELDS = [\n 'characterModel',\n 'locationModel',\n 'storyboardModel',\n 'editModel',\n] as const\n\nconst PROJECT_IMAGE_MODEL_FIELDS = [\n 'characterModel',\n 'locationModel',\n 'storyboardModel',\n 'editModel',\n] as const\n\ntype UserImageModelField = typeof USER_IMAGE_MODEL_FIELDS[number]\ntype ProjectImageModelField = typeof PROJECT_IMAGE_MODEL_FIELDS[number]\n\ninterface UserPreferenceRow {\n id: string\n userId: string\n imageResolution: string\n capabilityDefaults: string | null\n characterModel: string | null\n locationModel: string | null\n storyboardModel: string | null\n editModel: string | null\n}\n\ninterface ProjectRow {\n id: string\n projectId: string\n imageResolution: string\n videoResolution: string\n capabilityOverrides: string | null\n characterModel: string | null\n locationModel: string | null\n storyboardModel: string | null\n editModel: string | null\n videoModel: string | null\n}\n\ninterface MigrationSummary {\n mode: 'dry-run' | 'apply'\n userPreference: {\n scanned: number\n updated: number\n migratedImageResolution: number\n }\n novelPromotionProject: {\n scanned: number\n updated: number\n migratedImageResolution: number\n migratedVideoResolution: number\n }\n}\n\nfunction isRecord(value: unknown): value is Record {\n return !!value && typeof value === 'object' && !Array.isArray(value)\n}\n\nfunction isCapabilityValue(value: unknown): value is CapabilityValue {\n return typeof value === 'string' || typeof value === 'number' || typeof value === 'boolean'\n}\n\nfunction normalizeSelections(raw: unknown): CapabilitySelections {\n if (!isRecord(raw)) return {}\n\n const normalized: CapabilitySelections = {}\n for (const [modelKey, rawSelection] of Object.entries(raw)) {\n if (!isRecord(rawSelection)) continue\n\n const nextSelection: Record = {}\n for (const [field, value] of Object.entries(rawSelection)) {\n if (isCapabilityValue(value)) {\n nextSelection[field] = value\n }\n }\n\n normalized[modelKey] = nextSelection\n }\n\n return normalized\n}\n\nfunction parseSelections(raw: string | null): CapabilitySelections {\n if (!raw) return {}\n try {\n return normalizeSelections(JSON.parse(raw) as unknown)\n } catch {\n return {}\n }\n}\n\nfunction serializeSelections(selections: CapabilitySelections): string | null {\n if (Object.keys(selections).length === 0) return null\n return JSON.stringify(selections)\n}\n\nfunction getCapabilityResolutionOptions(\n modelType: 'image' | 'video',\n modelKey: string,\n): string[] {\n const parsed = parseModelKeyStrict(modelKey)\n if (!parsed) return []\n\n const capabilities = findBuiltinCapabilities(modelType, parsed.provider, parsed.modelId)\n const namespace = capabilities?.[modelType]\n if (!namespace || !isRecord(namespace)) return []\n\n const resolutionOptions = namespace.resolutionOptions\n if (!Array.isArray(resolutionOptions)) return []\n\n return resolutionOptions.filter((item): item is string => typeof item === 'string' && item.trim().length > 0)\n}\n\nfunction ensureModelResolutionSelection(input: {\n modelType: 'image' | 'video'\n modelKey: string\n resolution: string\n selections: CapabilitySelections\n}): boolean {\n const options = getCapabilityResolutionOptions(input.modelType, input.modelKey)\n if (options.length === 0) return false\n if (!options.includes(input.resolution)) return false\n\n const current = input.selections[input.modelKey]\n if (current && current.resolution !== undefined) {\n return false\n }\n\n input.selections[input.modelKey] = {\n ...(current || {}),\n resolution: input.resolution,\n }\n return true\n}\n\nfunction collectModelKeys(\n row: RowType,\n fields: readonly (keyof RowType)[],\n): string[] {\n const modelKeys: string[] = []\n for (const field of fields) {\n const value = row[field]\n if (typeof value !== 'string') continue\n const trimmed = value.trim()\n if (!trimmed) continue\n modelKeys.push(trimmed)\n }\n return modelKeys\n}\n\nasync function migrateUserPreferences(summary: MigrationSummary) {\n const rows = await prisma.userPreference.findMany({\n select: {\n id: true,\n userId: true,\n imageResolution: true,\n capabilityDefaults: true,\n characterModel: true,\n locationModel: true,\n storyboardModel: true,\n editModel: true,\n },\n }) as UserPreferenceRow[]\n\n summary.userPreference.scanned = rows.length\n\n for (const row of rows) {\n const nextSelections = parseSelections(row.capabilityDefaults)\n const modelKeys = collectModelKeys(row, USER_IMAGE_MODEL_FIELDS)\n let changed = false\n\n for (const modelKey of modelKeys) {\n if (ensureModelResolutionSelection({\n modelType: 'image',\n modelKey,\n resolution: row.imageResolution,\n selections: nextSelections,\n })) {\n changed = true\n summary.userPreference.migratedImageResolution += 1\n }\n }\n\n if (!changed) continue\n summary.userPreference.updated += 1\n\n if (APPLY) {\n await prisma.userPreference.update({\n where: { id: row.id },\n data: {\n capabilityDefaults: serializeSelections(nextSelections),\n },\n })\n }\n }\n}\n\nasync function migrateProjects(summary: MigrationSummary) {\n const rows = await prisma.novelPromotionProject.findMany({\n select: {\n id: true,\n projectId: true,\n imageResolution: true,\n videoResolution: true,\n capabilityOverrides: true,\n characterModel: true,\n locationModel: true,\n storyboardModel: true,\n editModel: true,\n videoModel: true,\n },\n }) as ProjectRow[]\n\n summary.novelPromotionProject.scanned = rows.length\n\n for (const row of rows) {\n const nextSelections = parseSelections(row.capabilityOverrides)\n const imageModelKeys = collectModelKeys(row, PROJECT_IMAGE_MODEL_FIELDS)\n let changed = false\n\n for (const modelKey of imageModelKeys) {\n if (ensureModelResolutionSelection({\n modelType: 'image',\n modelKey,\n resolution: row.imageResolution,\n selections: nextSelections,\n })) {\n changed = true\n summary.novelPromotionProject.migratedImageResolution += 1\n }\n }\n\n if (typeof row.videoModel === 'string' && row.videoModel.trim()) {\n if (ensureModelResolutionSelection({\n modelType: 'video',\n modelKey: row.videoModel.trim(),\n resolution: row.videoResolution,\n selections: nextSelections,\n })) {\n changed = true\n summary.novelPromotionProject.migratedVideoResolution += 1\n }\n }\n\n if (!changed) continue\n summary.novelPromotionProject.updated += 1\n\n if (APPLY) {\n await prisma.novelPromotionProject.update({\n where: { id: row.id },\n data: {\n capabilityOverrides: serializeSelections(nextSelections),\n },\n })\n }\n }\n}\n\nasync function main() {\n const summary: MigrationSummary = {\n mode: APPLY ? 'apply' : 'dry-run',\n userPreference: {\n scanned: 0,\n updated: 0,\n migratedImageResolution: 0,\n },\n novelPromotionProject: {\n scanned: 0,\n updated: 0,\n migratedImageResolution: 0,\n migratedVideoResolution: 0,\n },\n }\n\n await migrateUserPreferences(summary)\n await migrateProjects(summary)\n\n process.stdout.write(`${JSON.stringify(summary, null, 2)}\\n`)\n}\n\nmain()\n .catch((error: unknown) => {\n const message = error instanceof Error ? error.message : String(error)\n const missingColumn =\n message.includes('capabilityDefaults') || message.includes('capabilityOverrides')\n if (missingColumn && message.includes('does not exist')) {\n process.stderr.write(\n '[migrate-capability-selections] FAILED: required DB columns are missing. ' +\n 'Apply SQL migration `prisma/migrations/20260215_add_capability_selection_columns.sql` first.\\n',\n )\n } else {\n process.stderr.write(`[migrate-capability-selections] FAILED: ${message}\\n`)\n }\n process.exitCode = 1\n })\n .finally(async () => {\n await prisma.$disconnect()\n })\n"
},
{
"path": "scripts/migrations/migrate-custom-pricing-v2.ts",
"content": "import { prisma } from '@/lib/prisma'\n\nconst APPLY = process.argv.includes('--apply')\n\ntype PreferenceRow = {\n id: string\n userId: string\n customModels: string | null\n}\n\ntype MigrationSummary = {\n mode: 'dry-run' | 'apply'\n scanned: number\n updatedRows: number\n migratedModels: number\n skippedInvalidRows: number\n}\n\nfunction isRecord(value: unknown): value is Record {\n return !!value && typeof value === 'object' && !Array.isArray(value)\n}\n\nfunction parseCustomModels(raw: string | null): unknown[] | null {\n if (!raw) return []\n try {\n const parsed = JSON.parse(raw) as unknown\n if (!Array.isArray(parsed)) return null\n return parsed\n } catch {\n return null\n }\n}\n\nfunction migrateLegacyCustomPricing(raw: unknown): {\n changed: boolean\n next: unknown\n} {\n if (!isRecord(raw)) {\n return { changed: false, next: raw }\n }\n\n const hasLegacyInput = typeof raw.input === 'number' && Number.isFinite(raw.input) && raw.input >= 0\n const hasLegacyOutput = typeof raw.output === 'number' && Number.isFinite(raw.output) && raw.output >= 0\n if (!hasLegacyInput && !hasLegacyOutput) {\n return { changed: false, next: raw }\n }\n\n const llmRaw = isRecord(raw.llm) ? raw.llm : {}\n const llmInput = typeof llmRaw.inputPerMillion === 'number' && Number.isFinite(llmRaw.inputPerMillion) && llmRaw.inputPerMillion >= 0\n ? llmRaw.inputPerMillion\n : (hasLegacyInput ? raw.input as number : undefined)\n const llmOutput = typeof llmRaw.outputPerMillion === 'number' && Number.isFinite(llmRaw.outputPerMillion) && llmRaw.outputPerMillion >= 0\n ? llmRaw.outputPerMillion\n : (hasLegacyOutput ? raw.output as number : undefined)\n\n const nextPricing: Record = {}\n for (const [key, value] of Object.entries(raw)) {\n if (key === 'input' || key === 'output') continue\n nextPricing[key] = value\n }\n\n nextPricing.llm = {\n ...(llmInput !== undefined ? { inputPerMillion: llmInput } : {}),\n ...(llmOutput !== undefined ? { outputPerMillion: llmOutput } : {}),\n }\n\n return {\n changed: true,\n next: nextPricing,\n }\n}\n\nfunction migrateCustomModel(rawModel: unknown): { changed: boolean; next: unknown } {\n if (!isRecord(rawModel)) {\n return { changed: false, next: rawModel }\n }\n\n const migratedPricing = migrateLegacyCustomPricing(rawModel.customPricing)\n if (!migratedPricing.changed) {\n return { changed: false, next: rawModel }\n }\n\n return {\n changed: true,\n next: {\n ...rawModel,\n customPricing: migratedPricing.next,\n },\n }\n}\n\nasync function main() {\n const summary: MigrationSummary = {\n mode: APPLY ? 'apply' : 'dry-run',\n scanned: 0,\n updatedRows: 0,\n migratedModels: 0,\n skippedInvalidRows: 0,\n }\n\n const rows = await prisma.userPreference.findMany({\n select: {\n id: true,\n userId: true,\n customModels: true,\n },\n }) as PreferenceRow[]\n\n summary.scanned = rows.length\n\n for (const row of rows) {\n const parsedModels = parseCustomModels(row.customModels)\n if (parsedModels === null) {\n summary.skippedInvalidRows += 1\n continue\n }\n\n let rowChanged = false\n const nextModels = parsedModels.map((model) => {\n const migrated = migrateCustomModel(model)\n if (migrated.changed) {\n rowChanged = true\n summary.migratedModels += 1\n }\n return migrated.next\n })\n\n if (!rowChanged) continue\n summary.updatedRows += 1\n\n if (APPLY) {\n await prisma.userPreference.update({\n where: { id: row.id },\n data: {\n customModels: JSON.stringify(nextModels),\n },\n })\n }\n }\n\n console.log(JSON.stringify(summary, null, 2))\n}\n\nmain()\n .then(async () => {\n await prisma.$disconnect()\n })\n .catch(async (error: unknown) => {\n console.error('[migrate-custom-pricing-v2] failed', error)\n await prisma.$disconnect()\n process.exit(1)\n })\n"
},
{
"path": "scripts/migrations/migrate-gateway-route-openai-compat.ts",
"content": "import { prisma } from '@/lib/prisma'\nimport { migrateGatewayRoutePayload } from '@/lib/migrations/gateway-route-openai-compat'\n\nconst APPLY = process.argv.includes('--apply')\n\ntype PreferenceRow = {\n id: string\n userId: string\n customProviders: string | null\n}\n\ntype MigrationSummary = {\n mode: 'dry-run' | 'apply'\n scanned: number\n updatedRows: number\n migratedProviders: number\n routeLitellmToOpenaiCompat: number\n routeForcedOfficial: number\n geminiApiModeCorrected: number\n skippedInvalidRows: number\n}\n\nasync function main() {\n const summary: MigrationSummary = {\n mode: APPLY ? 'apply' : 'dry-run',\n scanned: 0,\n updatedRows: 0,\n migratedProviders: 0,\n routeLitellmToOpenaiCompat: 0,\n routeForcedOfficial: 0,\n geminiApiModeCorrected: 0,\n skippedInvalidRows: 0,\n }\n\n const rows = await prisma.userPreference.findMany({\n select: {\n id: true,\n userId: true,\n customProviders: true,\n },\n }) as PreferenceRow[]\n summary.scanned = rows.length\n\n for (const row of rows) {\n const result = migrateGatewayRoutePayload(row.customProviders)\n if (result.status === 'invalid') {\n summary.skippedInvalidRows += 1\n continue\n }\n\n summary.migratedProviders += result.summary.providersChanged\n summary.routeLitellmToOpenaiCompat += result.summary.routeLitellmToOpenaiCompat\n summary.routeForcedOfficial += result.summary.routeForcedOfficial\n summary.geminiApiModeCorrected += result.summary.geminiApiModeCorrected\n\n if (!result.changed) continue\n summary.updatedRows += 1\n\n if (APPLY) {\n await prisma.userPreference.update({\n where: { id: row.id },\n data: {\n customProviders: result.nextRaw ?? null,\n },\n })\n }\n }\n\n console.log(JSON.stringify(summary, null, 2))\n}\n\nmain()\n .then(async () => {\n await prisma.$disconnect()\n })\n .catch(async (error: unknown) => {\n console.error('[migrate-gateway-route-openai-compat] failed', error)\n await prisma.$disconnect()\n process.exit(1)\n })\n"
},
{
"path": "scripts/migrations/migrate-graph-artifacts-unique-index.ts",
"content": "import { prisma } from '@/lib/prisma'\n\nconst APPLY = process.argv.includes('--apply')\nconst REQUIRED_INDEX_NAME = 'graph_artifacts_runId_stepKey_artifactType_refId_key'\nconst REQUIRED_COLUMNS = ['runId', 'stepKey', 'artifactType', 'refId'] as const\n\ntype IndexRow = {\n Key_name: string\n Non_unique: number | string\n Seq_in_index: number | string\n Column_name: string\n}\n\ntype DuplicateRow = {\n runId: string\n stepKey: string\n artifactType: string\n refId: string\n c: bigint | number\n}\n\ntype MigrationSummary = {\n mode: 'dry-run' | 'apply'\n hasRequiredIndexBefore: boolean\n duplicateGroupCount: number\n duplicateSamples: Array<{\n runId: string\n stepKey: string\n artifactType: string\n refId: string\n count: number\n }>\n altered: boolean\n hasRequiredIndexAfter: boolean\n}\n\nfunction parseIntSafe(value: number | string) {\n if (typeof value === 'number') return value\n return Number.parseInt(value, 10)\n}\n\nfunction hasRequiredUniqueIndex(rows: IndexRow[]) {\n const grouped = new Map>()\n for (const row of rows) {\n const seq = parseIntSafe(row.Seq_in_index)\n const nonUnique = parseIntSafe(row.Non_unique)\n if (!Number.isFinite(seq) || !Number.isFinite(nonUnique)) continue\n const key = row.Key_name\n const items = grouped.get(key) || []\n items.push({\n seq,\n column: row.Column_name,\n nonUnique,\n })\n grouped.set(key, items)\n }\n\n for (const [key, entries] of grouped.entries()) {\n if (entries.length !== REQUIRED_COLUMNS.length) continue\n const sorted = entries.sort((a, b) => a.seq - b.seq)\n if (sorted[0]?.nonUnique !== 0) continue\n const columns = sorted.map((entry) => entry.column)\n const isTarget = columns.every((column, index) => column === REQUIRED_COLUMNS[index])\n if (isTarget && key === REQUIRED_INDEX_NAME) return true\n if (isTarget) return true\n }\n return false\n}\n\nfunction toNumber(value: bigint | number) {\n if (typeof value === 'bigint') return Number(value)\n return value\n}\n\nasync function loadIndexRows() {\n return await prisma.$queryRawUnsafe('SHOW INDEX FROM graph_artifacts')\n}\n\nasync function loadDuplicateGroups() {\n return await prisma.$queryRawUnsafe(\n `SELECT runId, stepKey, artifactType, refId, COUNT(*) AS c\n FROM graph_artifacts\n WHERE stepKey IS NOT NULL\n GROUP BY runId, stepKey, artifactType, refId\n HAVING c > 1\n LIMIT 20`,\n )\n}\n\nasync function main() {\n const beforeRows = await loadIndexRows()\n const hasBefore = hasRequiredUniqueIndex(beforeRows)\n const duplicates = await loadDuplicateGroups()\n\n const summary: MigrationSummary = {\n mode: APPLY ? 'apply' : 'dry-run',\n hasRequiredIndexBefore: hasBefore,\n duplicateGroupCount: duplicates.length,\n duplicateSamples: duplicates.map((row) => ({\n runId: row.runId,\n stepKey: row.stepKey,\n artifactType: row.artifactType,\n refId: row.refId,\n count: toNumber(row.c),\n })),\n altered: false,\n hasRequiredIndexAfter: hasBefore,\n }\n\n if (hasBefore) {\n console.log(JSON.stringify(summary, null, 2))\n return\n }\n\n if (duplicates.length > 0) {\n throw new Error(\n `cannot add unique index; found ${duplicates.length} duplicate groups in graph_artifacts (stepKey IS NOT NULL)`,\n )\n }\n\n if (APPLY) {\n await prisma.$executeRawUnsafe(\n `ALTER TABLE graph_artifacts\n ADD UNIQUE INDEX ${REQUIRED_INDEX_NAME} (runId, stepKey, artifactType, refId)`,\n )\n summary.altered = true\n const afterRows = await loadIndexRows()\n summary.hasRequiredIndexAfter = hasRequiredUniqueIndex(afterRows)\n if (!summary.hasRequiredIndexAfter) {\n throw new Error('unique index create verification failed')\n }\n }\n\n console.log(JSON.stringify(summary, null, 2))\n}\n\nmain()\n .then(async () => {\n await prisma.$disconnect()\n })\n .catch(async (error: unknown) => {\n console.error('[migrate-graph-artifacts-unique-index] failed', error)\n await prisma.$disconnect()\n process.exit(1)\n })\n"
},
{
"path": "scripts/migrations/migrate-model-config-contract.ts",
"content": "import fs from 'fs'\nimport path from 'path'\nimport { prisma } from '@/lib/prisma'\nimport {\n composeModelKey,\n parseModelKeyStrict,\n validateModelCapabilities,\n type ModelCapabilities,\n type UnifiedModelType,\n} from '@/lib/model-config-contract'\n\ntype ModelField =\n | 'analysisModel'\n | 'characterModel'\n | 'locationModel'\n | 'storyboardModel'\n | 'editModel'\n | 'videoModel'\n\ntype PreferenceRow = {\n id: string\n userId: string\n customModels: string | null\n analysisModel: string | null\n characterModel: string | null\n locationModel: string | null\n storyboardModel: string | null\n editModel: string | null\n videoModel: string | null\n}\n\ntype ProjectRow = {\n id: string\n projectId: string\n analysisModel: string | null\n characterModel: string | null\n locationModel: string | null\n storyboardModel: string | null\n editModel: string | null\n videoModel: string | null\n project: {\n userId: string\n }\n}\n\ntype MigrationIssue = {\n table: 'userPreference' | 'novelPromotionProject'\n rowId: string\n userId?: string\n field: string\n kind:\n | 'CUSTOM_MODELS_JSON_INVALID'\n | 'MODEL_SHAPE_INVALID'\n | 'MODEL_TYPE_INVALID'\n | 'MODEL_KEY_INCOMPLETE'\n | 'MODEL_KEY_MISMATCH'\n | 'MODEL_CAPABILITY_INVALID'\n | 'LEGACY_MODEL_ID_NOT_FOUND'\n | 'LEGACY_MODEL_ID_AMBIGUOUS'\n rawValue?: string | null\n candidates?: string[]\n message: string\n}\n\ntype MigrationReport = {\n generatedAt: string\n mode: 'dry-run' | 'apply'\n userPreference: {\n scanned: number\n updated: number\n updatedCustomModels: number\n updatedDefaultFields: number\n }\n novelPromotionProject: {\n scanned: number\n updated: number\n updatedFields: number\n }\n issues: MigrationIssue[]\n}\n\ntype NormalizedModel = {\n provider: string\n modelId: string\n modelKey: string\n name: string\n type: UnifiedModelType\n price: number\n resolution?: '2K' | '4K'\n capabilities?: ModelCapabilities\n}\n\nconst APPLY = process.argv.includes('--apply')\nconst MAX_ISSUES = 500\nconst MODEL_FIELDS: readonly ModelField[] = [\n 'analysisModel',\n 'characterModel',\n 'locationModel',\n 'storyboardModel',\n 'editModel',\n 'videoModel',\n]\n\nconst LEGACY_MODEL_ID_MAP = new Map([\n ['anthropic/claude-sonnet-4.5', 'openrouter::anthropic/claude-sonnet-4.5'],\n ['google/gemini-3-pro-preview', 'openrouter::google/gemini-3-pro-preview'],\n ['openai/gpt-5.2', 'openrouter::openai/gpt-5.2'],\n ['banana', 'fal::banana'],\n ['banana-2k', 'fal::banana'],\n ['seedream', 'ark::doubao-seedream-4-0-250828'],\n ['seedream4.5', 'ark::doubao-seedream-4-5-251128'],\n ['gemini-3-pro-image-preview', 'google::gemini-3-pro-image-preview'],\n ['gemini-3-pro-image-preview-batch', 'google::gemini-3-pro-image-preview-batch'],\n ['nano-banana-pro', 'google::gemini-3-pro-image-preview'],\n ['gemini-3.0-pro-image-portrait', 'flow2api::gemini-3.0-pro-image-portrait'],\n ['imagen-4.0-ultra-generate-001', 'google::imagen-4.0-ultra-generate-001'],\n ['doubao-seedance-1-0-pro-250528', 'ark::doubao-seedance-1-0-pro-250528'],\n ['doubao-seedance-1-0-pro-fast-251015', 'ark::doubao-seedance-1-0-pro-fast-251015'],\n ['doubao-seedance-1-0-pro-fast-251015-batch', 'ark::doubao-seedance-1-0-pro-fast-251015-batch'],\n])\n\nfunction parseReportPathArg(): string {\n const flagPrefix = '--report='\n const inline = process.argv.find((arg) => arg.startsWith(flagPrefix))\n if (inline) return inline.slice(flagPrefix.length)\n const flagIndex = process.argv.findIndex((arg) => arg === '--report')\n if (flagIndex !== -1 && process.argv[flagIndex + 1]) {\n return process.argv[flagIndex + 1]\n }\n return 'scripts/migrations/reports/model-config-migration-report.json'\n}\n\nfunction isRecord(value: unknown): value is Record {\n return !!value && typeof value === 'object' && !Array.isArray(value)\n}\n\nfunction toTrimmedString(value: unknown): string {\n return typeof value === 'string' ? value.trim() : ''\n}\n\nfunction isUnifiedModelType(value: unknown): value is UnifiedModelType {\n return value === 'llm'\n || value === 'image'\n || value === 'video'\n || value === 'audio'\n || value === 'lipsync'\n}\n\nfunction stableStringify(value: unknown): string {\n return JSON.stringify(value)\n}\n\nfunction parseCustomModels(raw: string | null): { ok: true; value: unknown[] } | { ok: false } {\n if (!raw) return { ok: true, value: [] }\n try {\n const parsed = JSON.parse(raw) as unknown\n if (!Array.isArray(parsed)) return { ok: false }\n return { ok: true, value: parsed }\n } catch {\n return { ok: false }\n }\n}\n\nfunction normalizeModel(\n raw: unknown,\n): { normalized: NormalizedModel | null; changed: boolean; issue?: Omit } {\n if (!isRecord(raw)) {\n return {\n normalized: null,\n changed: false,\n issue: {\n field: 'customModels',\n kind: 'MODEL_SHAPE_INVALID',\n message: 'customModels item must be object',\n },\n }\n }\n\n const modelType = raw.type\n if (!isUnifiedModelType(modelType)) {\n return {\n normalized: null,\n changed: false,\n issue: {\n field: 'customModels.type',\n kind: 'MODEL_TYPE_INVALID',\n rawValue: String(raw.type ?? ''),\n message: 'custom model type must be llm/image/video/audio/lipsync',\n },\n }\n }\n\n const providerField = toTrimmedString(raw.provider)\n const modelIdField = toTrimmedString(raw.modelId)\n const parsedModelKey = parseModelKeyStrict(toTrimmedString(raw.modelKey))\n\n const provider = providerField || parsedModelKey?.provider || ''\n const modelId = modelIdField || parsedModelKey?.modelId || ''\n const modelKey = composeModelKey(provider, modelId)\n if (!modelKey) {\n return {\n normalized: null,\n changed: false,\n issue: {\n field: 'customModels.modelKey',\n kind: 'MODEL_KEY_INCOMPLETE',\n rawValue: toTrimmedString(raw.modelKey),\n message: 'provider/modelId/modelKey cannot compose a valid model_key',\n },\n }\n }\n\n if (parsedModelKey && parsedModelKey.modelKey !== modelKey) {\n return {\n normalized: null,\n changed: false,\n issue: {\n field: 'customModels.modelKey',\n kind: 'MODEL_KEY_MISMATCH',\n rawValue: toTrimmedString(raw.modelKey),\n message: 'modelKey conflicts with provider/modelId',\n },\n }\n }\n\n const rawResolution = toTrimmedString(raw.resolution)\n const resolution = rawResolution === '2K' || rawResolution === '4K' ? rawResolution : undefined\n const capabilities = isRecord(raw.capabilities)\n ? ({ ...(raw.capabilities as ModelCapabilities) })\n : undefined\n const capabilityIssues = validateModelCapabilities(modelType, capabilities)\n if (capabilityIssues.length > 0) {\n const firstIssue = capabilityIssues[0]\n return {\n normalized: null,\n changed: false,\n issue: {\n field: firstIssue.field,\n kind: 'MODEL_CAPABILITY_INVALID',\n message: `${firstIssue.code}: ${firstIssue.message}`,\n },\n }\n }\n\n const name = toTrimmedString(raw.name) || modelId\n const price = typeof raw.price === 'number' && Number.isFinite(raw.price) ? raw.price : 0\n\n const normalized: NormalizedModel = {\n provider,\n modelId,\n modelKey,\n name,\n type: modelType,\n price,\n ...(resolution ? { resolution } : {}),\n ...(capabilities ? { capabilities } : {}),\n }\n\n const changed = stableStringify(raw) !== stableStringify(normalized)\n return { normalized, changed }\n}\n\nfunction addIssue(report: MigrationReport, issue: MigrationIssue) {\n if (report.issues.length >= MAX_ISSUES) return\n report.issues.push(issue)\n}\n\nfunction normalizeModelFieldValue(\n rawValue: string | null,\n field: ModelField,\n mappingByModelId: Map,\n): { nextValue: string | null; changed: boolean; issue?: Omit } {\n if (!rawValue || !rawValue.trim()) return { nextValue: null, changed: rawValue !== null }\n const trimmed = rawValue.trim()\n const parsed = parseModelKeyStrict(trimmed)\n if (parsed) {\n return { nextValue: parsed.modelKey, changed: parsed.modelKey !== rawValue }\n }\n\n const candidates = mappingByModelId.get(trimmed) || []\n if (candidates.length === 1) {\n return { nextValue: candidates[0], changed: candidates[0] !== rawValue }\n }\n if (candidates.length === 0) {\n const mappedModelKey = LEGACY_MODEL_ID_MAP.get(trimmed)\n if (mappedModelKey) {\n return { nextValue: mappedModelKey, changed: mappedModelKey !== rawValue }\n }\n }\n if (candidates.length === 0) {\n return {\n nextValue: rawValue,\n changed: false,\n issue: {\n field,\n kind: 'LEGACY_MODEL_ID_NOT_FOUND',\n rawValue,\n message: `${field} legacy modelId cannot be mapped`,\n },\n }\n }\n return {\n nextValue: rawValue,\n changed: false,\n issue: {\n field,\n kind: 'LEGACY_MODEL_ID_AMBIGUOUS',\n rawValue,\n candidates,\n message: `${field} legacy modelId maps to multiple providers`,\n },\n }\n}\n\nasync function main() {\n const reportPath = parseReportPathArg()\n const report: MigrationReport = {\n generatedAt: new Date().toISOString(),\n mode: APPLY ? 'apply' : 'dry-run',\n userPreference: {\n scanned: 0,\n updated: 0,\n updatedCustomModels: 0,\n updatedDefaultFields: 0,\n },\n novelPromotionProject: {\n scanned: 0,\n updated: 0,\n updatedFields: 0,\n },\n issues: [],\n }\n\n const userPrefs = await prisma.userPreference.findMany({\n select: {\n id: true,\n userId: true,\n customModels: true,\n analysisModel: true,\n characterModel: true,\n locationModel: true,\n storyboardModel: true,\n editModel: true,\n videoModel: true,\n },\n })\n\n const userMappings = new Map>()\n\n for (const pref of userPrefs) {\n report.userPreference.scanned += 1\n const updateData: Partial> = {}\n\n const parsedCustomModels = parseCustomModels(pref.customModels)\n const normalizedModels: NormalizedModel[] = []\n let customModelsChanged = false\n\n if (!parsedCustomModels.ok) {\n addIssue(report, {\n table: 'userPreference',\n rowId: pref.id,\n userId: pref.userId,\n field: 'customModels',\n kind: 'CUSTOM_MODELS_JSON_INVALID',\n rawValue: pref.customModels,\n message: 'customModels JSON is invalid',\n })\n } else {\n for (let index = 0; index < parsedCustomModels.value.length; index += 1) {\n const normalizedResult = normalizeModel(parsedCustomModels.value[index])\n if (normalizedResult.issue) {\n addIssue(report, {\n table: 'userPreference',\n rowId: pref.id,\n userId: pref.userId,\n ...normalizedResult.issue,\n })\n continue\n }\n if (normalizedResult.normalized) {\n normalizedModels.push(normalizedResult.normalized)\n if (normalizedResult.changed) customModelsChanged = true\n }\n }\n }\n\n const mappingByModelId = new Map()\n for (const model of normalizedModels) {\n const existing = mappingByModelId.get(model.modelId) || []\n if (!existing.includes(model.modelKey)) existing.push(model.modelKey)\n mappingByModelId.set(model.modelId, existing)\n }\n userMappings.set(pref.userId, mappingByModelId)\n\n if (customModelsChanged) {\n updateData.customModels = JSON.stringify(normalizedModels)\n report.userPreference.updatedCustomModels += 1\n }\n\n for (const field of MODEL_FIELDS) {\n const normalizedField = normalizeModelFieldValue(pref[field], field, mappingByModelId)\n if (normalizedField.issue) {\n addIssue(report, {\n table: 'userPreference',\n rowId: pref.id,\n userId: pref.userId,\n ...normalizedField.issue,\n })\n }\n if (normalizedField.changed) {\n updateData[field] = normalizedField.nextValue\n report.userPreference.updatedDefaultFields += 1\n }\n }\n\n if (Object.keys(updateData).length > 0) {\n report.userPreference.updated += 1\n if (APPLY) {\n await prisma.userPreference.update({\n where: { id: pref.id },\n data: updateData,\n })\n }\n }\n }\n\n const projects = await prisma.novelPromotionProject.findMany({\n select: {\n id: true,\n projectId: true,\n analysisModel: true,\n characterModel: true,\n locationModel: true,\n storyboardModel: true,\n editModel: true,\n videoModel: true,\n project: {\n select: {\n userId: true,\n },\n },\n },\n })\n\n for (const row of projects as ProjectRow[]) {\n report.novelPromotionProject.scanned += 1\n const mappingByModelId = userMappings.get(row.project.userId) || new Map()\n const updateData: Partial> = {}\n\n for (const field of MODEL_FIELDS) {\n const normalizedField = normalizeModelFieldValue(row[field], field, mappingByModelId)\n if (normalizedField.issue) {\n addIssue(report, {\n table: 'novelPromotionProject',\n rowId: row.id,\n userId: row.project.userId,\n ...normalizedField.issue,\n })\n }\n if (normalizedField.changed) {\n updateData[field] = normalizedField.nextValue\n report.novelPromotionProject.updatedFields += 1\n }\n }\n\n if (Object.keys(updateData).length > 0) {\n report.novelPromotionProject.updated += 1\n if (APPLY) {\n await prisma.novelPromotionProject.update({\n where: { id: row.id },\n data: updateData,\n })\n }\n }\n }\n\n const absoluteReportPath = path.isAbsolute(reportPath)\n ? reportPath\n : path.join(process.cwd(), reportPath)\n fs.mkdirSync(path.dirname(absoluteReportPath), { recursive: true })\n fs.writeFileSync(absoluteReportPath, `${JSON.stringify(report, null, 2)}\\n`, 'utf8')\n\n process.stdout.write(\n `[migrate-model-config-contract] mode=${report.mode} ` +\n `prefs=${report.userPreference.scanned}/${report.userPreference.updated} ` +\n `projects=${report.novelPromotionProject.scanned}/${report.novelPromotionProject.updated} ` +\n `issues=${report.issues.length} report=${absoluteReportPath}\\n`,\n )\n}\n\nmain()\n .catch((error) => {\n process.stderr.write(`[migrate-model-config-contract] failed: ${String(error)}\\n`)\n process.exitCode = 1\n })\n .finally(async () => {\n await prisma.$disconnect()\n })\n"
},
{
"path": "scripts/migrations/migrate-qwen-to-bailian.ts",
"content": "import { prisma } from '@/lib/prisma'\nimport { composeModelKey, parseModelKeyStrict, type CapabilitySelections } from '@/lib/model-config-contract'\n\nconst APPLY = process.argv.includes('--apply')\n\ntype PreferenceRow = {\n id: string\n userId: string\n customProviders: string | null\n customModels: string | null\n analysisModel: string | null\n characterModel: string | null\n locationModel: string | null\n storyboardModel: string | null\n editModel: string | null\n videoModel: string | null\n lipSyncModel: string | null\n capabilityDefaults: string | null\n}\n\ntype StoredProvider = {\n id: string\n name: string\n baseUrl?: string\n apiKey?: string\n apiMode?: 'gemini-sdk' | 'openai-official'\n gatewayRoute?: 'official' | 'litellm'\n}\n\ntype StoredModel = {\n modelId: string\n modelKey: string\n name: string\n type: string\n provider: string\n price: number\n}\n\ntype MigrationConflict = {\n userId: string\n reason: string\n}\n\ntype MigrationSummary = {\n mode: 'dry-run' | 'apply'\n scanned: number\n updatedRows: number\n updatedProviders: number\n updatedModels: number\n updatedDefaults: number\n updatedCapabilityDefaults: number\n invalidRows: number\n conflicts: MigrationConflict[]\n}\n\ntype DefaultModelField =\n | 'analysisModel'\n | 'characterModel'\n | 'locationModel'\n | 'storyboardModel'\n | 'editModel'\n | 'videoModel'\n | 'lipSyncModel'\n\nconst DEFAULT_MODEL_FIELDS: readonly DefaultModelField[] = [\n 'analysisModel',\n 'characterModel',\n 'locationModel',\n 'storyboardModel',\n 'editModel',\n 'videoModel',\n 'lipSyncModel',\n]\n\nfunction isRecord(value: unknown): value is Record {\n return !!value && typeof value === 'object' && !Array.isArray(value)\n}\n\nfunction readTrimmedString(value: unknown): string {\n return typeof value === 'string' ? value.trim() : ''\n}\n\nfunction parseProviders(raw: string | null): StoredProvider[] | null {\n if (!raw) return []\n try {\n const parsed = JSON.parse(raw) as unknown\n if (!Array.isArray(parsed)) return null\n const providers: StoredProvider[] = []\n for (const item of parsed) {\n if (!isRecord(item)) return null\n const id = readTrimmedString(item.id)\n const name = readTrimmedString(item.name)\n if (!id || !name) return null\n const provider: StoredProvider = { id, name }\n if (typeof item.baseUrl === 'string' && item.baseUrl.trim()) provider.baseUrl = item.baseUrl.trim()\n if (typeof item.apiKey === 'string' && item.apiKey.trim()) provider.apiKey = item.apiKey.trim()\n if (item.apiMode === 'gemini-sdk' || item.apiMode === 'openai-official') provider.apiMode = item.apiMode\n if (item.gatewayRoute === 'official' || item.gatewayRoute === 'litellm') provider.gatewayRoute = item.gatewayRoute\n providers.push(provider)\n }\n return providers\n } catch {\n return null\n }\n}\n\nfunction parseModels(raw: string | null): StoredModel[] | null {\n if (!raw) return []\n try {\n const parsed = JSON.parse(raw) as unknown\n if (!Array.isArray(parsed)) return null\n const models: StoredModel[] = []\n for (const item of parsed) {\n if (!isRecord(item)) return null\n const modelId = readTrimmedString(item.modelId)\n const modelKey = readTrimmedString(item.modelKey)\n const provider = readTrimmedString(item.provider)\n const name = readTrimmedString(item.name)\n const type = readTrimmedString(item.type)\n const price = typeof item.price === 'number' && Number.isFinite(item.price) ? item.price : 0\n if (!modelId || !provider || !type) return null\n const normalizedModelKey = modelKey || composeModelKey(provider, modelId)\n if (!normalizedModelKey) return null\n models.push({\n modelId,\n modelKey: normalizedModelKey,\n provider,\n name: name || modelId,\n type,\n price,\n })\n }\n return models\n } catch {\n return null\n }\n}\n\nfunction parseCapabilityDefaults(raw: string | null): CapabilitySelections | null {\n if (!raw) return {}\n try {\n const parsed = JSON.parse(raw) as unknown\n if (!isRecord(parsed)) return null\n const selections: CapabilitySelections = {}\n for (const [modelKey, value] of Object.entries(parsed)) {\n if (!isRecord(value)) continue\n const nextSelection: Record = {}\n for (const [field, option] of Object.entries(value)) {\n if (typeof option === 'string' || typeof option === 'number' || typeof option === 'boolean') {\n nextSelection[field] = option\n }\n }\n selections[modelKey] = nextSelection\n }\n return selections\n } catch {\n return null\n }\n}\n\nfunction migrateProviderId(providerId: string): string {\n if (providerId === 'qwen') return 'bailian'\n const parsed = parseModelKeyStrict(providerId)\n if (parsed) return providerId\n const marker = providerId.indexOf(':')\n if (marker === -1) return providerId\n const providerKey = providerId.slice(0, marker)\n if (providerKey !== 'qwen') return providerId\n return `bailian${providerId.slice(marker)}`\n}\n\nfunction migrateModelKey(rawModelKey: string): string {\n const parsed = parseModelKeyStrict(rawModelKey)\n if (!parsed) return rawModelKey\n if (parsed.provider !== 'qwen') return parsed.modelKey\n return composeModelKey('bailian', parsed.modelId)\n}\n\nfunction migrateDefaultModel(rawValue: string | null): string | null {\n if (!rawValue) return rawValue\n const value = rawValue.trim()\n if (!value) return null\n return migrateModelKey(value)\n}\n\nfunction hasProviderByKey(providers: StoredProvider[], providerKey: string): boolean {\n return providers.some((provider) => {\n const marker = provider.id.indexOf(':')\n const key = marker === -1 ? provider.id : provider.id.slice(0, marker)\n return key === providerKey\n })\n}\n\nasync function main() {\n const summary: MigrationSummary = {\n mode: APPLY ? 'apply' : 'dry-run',\n scanned: 0,\n updatedRows: 0,\n updatedProviders: 0,\n updatedModels: 0,\n updatedDefaults: 0,\n updatedCapabilityDefaults: 0,\n invalidRows: 0,\n conflicts: [],\n }\n\n const rows = await prisma.userPreference.findMany({\n select: {\n id: true,\n userId: true,\n customProviders: true,\n customModels: true,\n analysisModel: true,\n characterModel: true,\n locationModel: true,\n storyboardModel: true,\n editModel: true,\n videoModel: true,\n lipSyncModel: true,\n capabilityDefaults: true,\n },\n }) as PreferenceRow[]\n\n summary.scanned = rows.length\n\n for (const row of rows) {\n const providers = parseProviders(row.customProviders)\n const models = parseModels(row.customModels)\n const capabilityDefaults = parseCapabilityDefaults(row.capabilityDefaults)\n if (!providers || !models || !capabilityDefaults) {\n summary.invalidRows += 1\n continue\n }\n\n const hasQwenProvider = hasProviderByKey(providers, 'qwen')\n const hasBailianProvider = hasProviderByKey(providers, 'bailian')\n if (hasQwenProvider && hasBailianProvider) {\n summary.conflicts.push({\n userId: row.userId,\n reason: 'both qwen and bailian providers exist',\n })\n continue\n }\n\n let rowChanged = false\n\n const nextProviders = providers.map((provider) => {\n const nextId = migrateProviderId(provider.id)\n if (nextId !== provider.id) {\n rowChanged = true\n summary.updatedProviders += 1\n }\n return {\n ...provider,\n id: nextId,\n ...(nextId === 'bailian' ? { name: 'Alibaba Bailian' } : {}),\n }\n })\n\n const nextModels = models.map((model) => {\n const nextProvider = migrateProviderId(model.provider)\n const nextModelKey = migrateModelKey(model.modelKey)\n const changed = nextProvider !== model.provider || nextModelKey !== model.modelKey\n if (changed) {\n rowChanged = true\n summary.updatedModels += 1\n }\n return {\n ...model,\n provider: nextProvider,\n modelKey: nextModelKey,\n }\n })\n const modelKeySet = new Set()\n let hasModelConflict = false\n for (const model of nextModels) {\n if (!modelKeySet.has(model.modelKey)) {\n modelKeySet.add(model.modelKey)\n continue\n }\n hasModelConflict = true\n break\n }\n if (hasModelConflict) {\n summary.conflicts.push({\n userId: row.userId,\n reason: 'model key collision after qwen -> bailian migration',\n })\n continue\n }\n\n const nextDefaults: Partial> = {}\n for (const field of DEFAULT_MODEL_FIELDS) {\n const current = row[field]\n const next = migrateDefaultModel(current)\n nextDefaults[field] = next\n if ((current || null) !== (next || null)) {\n rowChanged = true\n summary.updatedDefaults += 1\n }\n }\n\n const nextCapabilityDefaults: CapabilitySelections = {}\n for (const [modelKey, selection] of Object.entries(capabilityDefaults)) {\n const nextModelKey = migrateModelKey(modelKey)\n nextCapabilityDefaults[nextModelKey] = selection\n if (nextModelKey !== modelKey) {\n rowChanged = true\n summary.updatedCapabilityDefaults += 1\n }\n }\n\n if (!rowChanged) continue\n summary.updatedRows += 1\n\n if (APPLY) {\n await prisma.userPreference.update({\n where: { id: row.id },\n data: {\n customProviders: JSON.stringify(nextProviders),\n customModels: JSON.stringify(nextModels),\n analysisModel: nextDefaults.analysisModel || null,\n characterModel: nextDefaults.characterModel || null,\n locationModel: nextDefaults.locationModel || null,\n storyboardModel: nextDefaults.storyboardModel || null,\n editModel: nextDefaults.editModel || null,\n videoModel: nextDefaults.videoModel || null,\n lipSyncModel: nextDefaults.lipSyncModel || null,\n capabilityDefaults: Object.keys(nextCapabilityDefaults).length > 0\n ? JSON.stringify(nextCapabilityDefaults)\n : null,\n },\n })\n }\n }\n\n console.log(JSON.stringify(summary, null, 2))\n if (summary.conflicts.length > 0) {\n process.exitCode = 2\n }\n}\n\nmain()\n .then(async () => {\n await prisma.$disconnect()\n })\n .catch(async (error: unknown) => {\n console.error('[migrate-qwen-to-bailian] failed', error)\n await prisma.$disconnect()\n process.exit(1)\n })\n"
},
{
"path": "scripts/migrations/migrate-release-blockers.ts",
"content": "import { prisma } from '@/lib/prisma'\nimport { composeModelKey, parseModelKeyStrict, type CapabilitySelections } from '@/lib/model-config-contract'\n\ntype Mode = 'dry-run' | 'apply'\n\ntype UserPreferenceRow = {\n id: string\n userId: string\n customProviders: string | null\n customModels: string | null\n analysisModel: string | null\n characterModel: string | null\n locationModel: string | null\n storyboardModel: string | null\n editModel: string | null\n videoModel: string | null\n audioModel: string | null\n lipSyncModel: string | null\n capabilityDefaults: string | null\n}\n\ntype NovelProjectRow = {\n id: string\n projectId: string\n analysisModel: string | null\n characterModel: string | null\n locationModel: string | null\n storyboardModel: string | null\n editModel: string | null\n videoModel: string | null\n capabilityOverrides: string | null\n}\n\ntype StoredProvider = {\n id: string\n name: string\n baseUrl?: string\n apiKey?: string\n apiMode?: 'gemini-sdk' | 'openai-official'\n gatewayRoute?: 'official' | 'openai-compat'\n}\n\ntype StoredModel = {\n modelId: string\n modelKey: string\n provider: string\n [key: string]: unknown\n}\n\ntype ParseResult = {\n ok: boolean\n value: T\n}\n\ntype MigrationSummary = {\n mode: Mode\n userPreference: {\n scanned: number\n updated: number\n dirtyClearedProviders: number\n dirtyClearedModels: number\n dirtyClearedCapabilityDefaults: number\n migratedProviders: number\n migratedModels: number\n migratedDefaultModelFields: number\n migratedCapabilityDefaultKeys: number\n modelCollisionsResolvedByBailian: number\n providerCollisionsResolvedByBailian: number\n invalidModelFieldsCleared: number\n }\n novelPromotionProject: {\n scanned: number\n updated: number\n migratedModelFields: number\n migratedCapabilityOverrideKeys: number\n invalidModelFieldsCleared: number\n dirtyClearedCapabilityOverrides: number\n }\n graphArtifacts: {\n hasRequiredUniqueIndexBefore: boolean\n duplicateGroupsBefore: number\n duplicateGroupSamples: Array<{\n runId: string\n stepKey: string\n artifactType: string\n refId: string\n count: number\n }>\n deletedRowsForDedup: number\n duplicateGroupsAfter: number\n indexAdded: boolean\n hasRequiredUniqueIndexAfter: boolean\n }\n}\n\ntype MysqlIndexRow = {\n Key_name: string\n Non_unique: number | string\n Seq_in_index: number | string\n Column_name: string\n}\n\ntype DuplicateGroupRow = {\n runId: string\n stepKey: string\n artifactType: string\n refId: string\n c: bigint | number\n}\n\ntype CountRow = {\n c: bigint | number\n}\n\ntype DefaultModelField =\n | 'analysisModel'\n | 'characterModel'\n | 'locationModel'\n | 'storyboardModel'\n | 'editModel'\n | 'videoModel'\n | 'audioModel'\n | 'lipSyncModel'\n\ntype ProjectModelField =\n | 'analysisModel'\n | 'characterModel'\n | 'locationModel'\n | 'storyboardModel'\n | 'editModel'\n | 'videoModel'\n\ntype UserPreferenceUpdateData = Partial> & {\n customProviders?: string | null\n customModels?: string | null\n capabilityDefaults?: string | null\n}\n\ntype NovelProjectUpdateData = Partial> & {\n capabilityOverrides?: string | null\n}\n\nconst MODE: Mode = process.argv.includes('--dry-run') ? 'dry-run' : 'apply'\nconst APPLY = MODE === 'apply'\n\nconst OFFICIAL_ONLY_PROVIDER_KEYS = new Set(['bailian', 'siliconflow'])\nconst DEFAULT_MODEL_FIELDS: readonly DefaultModelField[] = [\n 'analysisModel',\n 'characterModel',\n 'locationModel',\n 'storyboardModel',\n 'editModel',\n 'videoModel',\n 'audioModel',\n 'lipSyncModel',\n]\nconst PROJECT_MODEL_FIELDS: readonly ProjectModelField[] = [\n 'analysisModel',\n 'characterModel',\n 'locationModel',\n 'storyboardModel',\n 'editModel',\n 'videoModel',\n]\nconst REQUIRED_GRAPH_ARTIFACT_UNIQUE_COLUMNS = ['runId', 'stepKey', 'artifactType', 'refId'] as const\n\nfunction isRecord(value: unknown): value is Record {\n return !!value && typeof value === 'object' && !Array.isArray(value)\n}\n\nfunction readTrimmedString(value: unknown): string {\n return typeof value === 'string' ? value.trim() : ''\n}\n\nfunction toNullableModelField(raw: string | null | undefined): string | null {\n const trimmed = readTrimmedString(raw)\n return trimmed || null\n}\n\nfunction getProviderKey(providerId: string): string {\n const index = providerId.indexOf(':')\n return index === -1 ? providerId : providerId.slice(0, index)\n}\n\nfunction migrateProviderId(providerId: string): string {\n const trimmed = providerId.trim()\n if (!trimmed) return trimmed\n if (trimmed === 'qwen') return 'bailian'\n\n const providerKey = getProviderKey(trimmed)\n if (providerKey !== 'qwen') return trimmed\n return `bailian${trimmed.slice(providerKey.length)}`\n}\n\nfunction migrateModelKey(rawModelKey: string): string {\n const parsed = parseModelKeyStrict(rawModelKey)\n if (!parsed) return rawModelKey\n if (getProviderKey(parsed.provider) !== 'qwen') return parsed.modelKey\n const nextProvider = migrateProviderId(parsed.provider)\n return composeModelKey(nextProvider, parsed.modelId)\n}\n\nfunction providerPriorityByOriginalKey(originalProviderId: string): number {\n const key = getProviderKey(originalProviderId)\n if (key === 'bailian') return 2\n if (key === 'qwen') return 1\n return 0\n}\n\nfunction normalizeGatewayRoute(\n providerId: string,\n rawGatewayRoute: unknown,\n): 'official' | 'openai-compat' {\n const providerKey = getProviderKey(providerId)\n if (providerKey === 'openai-compatible') return 'openai-compat'\n if (providerKey === 'gemini-compatible') return 'official'\n if (OFFICIAL_ONLY_PROVIDER_KEYS.has(providerKey)) return 'official'\n return rawGatewayRoute === 'openai-compat' ? 'openai-compat' : 'official'\n}\n\nfunction parseJsonArray(raw: string | null): ParseResult {\n if (!raw) return { ok: true, value: [] }\n try {\n const parsed = JSON.parse(raw) as unknown\n if (!Array.isArray(parsed)) return { ok: false, value: [] }\n return { ok: true, value: parsed }\n } catch {\n return { ok: false, value: [] }\n }\n}\n\nfunction parseJsonRecord(raw: string | null): ParseResult> {\n if (!raw) return { ok: true, value: {} }\n try {\n const parsed = JSON.parse(raw) as unknown\n if (!isRecord(parsed)) return { ok: false, value: {} }\n return { ok: true, value: parsed }\n } catch {\n return { ok: false, value: {} }\n }\n}\n\nfunction migrateProviders(\n rawProviders: string | null,\n): {\n ok: boolean\n nextRaw: string | null\n changed: boolean\n migratedProviders: number\n collisionsResolvedByBailian: number\n} {\n const parsed = parseJsonArray(rawProviders)\n if (!parsed.ok) {\n return {\n ok: false,\n nextRaw: null,\n changed: rawProviders !== null,\n migratedProviders: 0,\n collisionsResolvedByBailian: 0,\n }\n }\n\n const deduped = new Map()\n let migratedProviders = 0\n let collisionsResolvedByBailian = 0\n\n for (const item of parsed.value) {\n if (!isRecord(item)) {\n return {\n ok: false,\n nextRaw: null,\n changed: true,\n migratedProviders: 0,\n collisionsResolvedByBailian: 0,\n }\n }\n\n const id = readTrimmedString(item.id)\n const name = readTrimmedString(item.name)\n if (!id || !name) {\n return {\n ok: false,\n nextRaw: null,\n changed: true,\n migratedProviders: 0,\n collisionsResolvedByBailian: 0,\n }\n }\n\n const nextId = migrateProviderId(id)\n if (nextId !== id) migratedProviders += 1\n\n const apiModeRaw = readTrimmedString(item.apiMode)\n let apiMode: 'gemini-sdk' | 'openai-official' | undefined\n if (apiModeRaw === 'gemini-sdk' || apiModeRaw === 'openai-official') {\n apiMode = apiModeRaw\n }\n if (getProviderKey(nextId) === 'gemini-compatible' && apiMode === 'openai-official') {\n apiMode = 'gemini-sdk'\n }\n\n const nextProvider: StoredProvider = {\n id: nextId,\n name: getProviderKey(nextId) === 'bailian' ? 'Alibaba Bailian' : name,\n baseUrl: readTrimmedString(item.baseUrl) || undefined,\n apiKey: typeof item.apiKey === 'string' ? item.apiKey.trim() : undefined,\n apiMode,\n gatewayRoute: normalizeGatewayRoute(nextId, item.gatewayRoute),\n }\n\n const dedupeKey = nextProvider.id.toLowerCase()\n const nextPriority = providerPriorityByOriginalKey(id)\n const existing = deduped.get(dedupeKey)\n if (!existing) {\n deduped.set(dedupeKey, { provider: nextProvider, priority: nextPriority })\n continue\n }\n\n if (nextPriority > existing.priority) {\n deduped.set(dedupeKey, { provider: nextProvider, priority: nextPriority })\n collisionsResolvedByBailian += 1\n }\n }\n\n const nextProviders = Array.from(deduped.values()).map((entry) => entry.provider)\n const nextRaw = nextProviders.length > 0 ? JSON.stringify(nextProviders) : null\n return {\n ok: true,\n nextRaw,\n changed: (rawProviders || null) !== (nextRaw || null),\n migratedProviders,\n collisionsResolvedByBailian,\n }\n}\n\nfunction migrateModels(\n rawModels: string | null,\n): {\n ok: boolean\n nextRaw: string | null\n changed: boolean\n migratedModels: number\n collisionsResolvedByBailian: number\n} {\n const parsed = parseJsonArray(rawModels)\n if (!parsed.ok) {\n return {\n ok: false,\n nextRaw: null,\n changed: rawModels !== null,\n migratedModels: 0,\n collisionsResolvedByBailian: 0,\n }\n }\n\n const deduped = new Map()\n let migratedModels = 0\n let collisionsResolvedByBailian = 0\n\n for (const item of parsed.value) {\n if (!isRecord(item)) {\n return {\n ok: false,\n nextRaw: null,\n changed: true,\n migratedModels: 0,\n collisionsResolvedByBailian: 0,\n }\n }\n\n const providerRaw = readTrimmedString(item.provider)\n const modelIdRaw = readTrimmedString(item.modelId)\n const modelKeyRaw = readTrimmedString(item.modelKey)\n const parsedModelKey = parseModelKeyStrict(modelKeyRaw)\n\n const sourceProvider = providerRaw || parsedModelKey?.provider || ''\n const sourceModelId = modelIdRaw || parsedModelKey?.modelId || ''\n if (!sourceProvider || !sourceModelId) {\n return {\n ok: false,\n nextRaw: null,\n changed: true,\n migratedModels: 0,\n collisionsResolvedByBailian: 0,\n }\n }\n\n const nextProvider = migrateProviderId(sourceProvider)\n const nextModelKey = composeModelKey(nextProvider, sourceModelId)\n if (!nextModelKey) {\n return {\n ok: false,\n nextRaw: null,\n changed: true,\n migratedModels: 0,\n collisionsResolvedByBailian: 0,\n }\n }\n\n if (nextProvider !== sourceProvider || nextModelKey !== modelKeyRaw) migratedModels += 1\n\n const nextModel: StoredModel = {\n ...item,\n provider: nextProvider,\n modelId: sourceModelId,\n modelKey: nextModelKey,\n }\n const dedupeKey = nextModelKey.toLowerCase()\n const nextPriority = providerPriorityByOriginalKey(sourceProvider)\n const existing = deduped.get(dedupeKey)\n if (!existing) {\n deduped.set(dedupeKey, { model: nextModel, priority: nextPriority })\n continue\n }\n\n if (nextPriority > existing.priority) {\n deduped.set(dedupeKey, { model: nextModel, priority: nextPriority })\n collisionsResolvedByBailian += 1\n }\n }\n\n const nextModels = Array.from(deduped.values()).map((entry) => entry.model)\n const nextRaw = nextModels.length > 0 ? JSON.stringify(nextModels) : null\n return {\n ok: true,\n nextRaw,\n changed: (rawModels || null) !== (nextRaw || null),\n migratedModels,\n collisionsResolvedByBailian,\n }\n}\n\nfunction migrateModelField(\n raw: string | null,\n): {\n nextValue: string | null\n changed: boolean\n migrated: boolean\n clearedInvalid: boolean\n} {\n const current = toNullableModelField(raw)\n if (!current) {\n return {\n nextValue: null,\n changed: current !== raw,\n migrated: false,\n clearedInvalid: false,\n }\n }\n\n const parsed = parseModelKeyStrict(current)\n if (!parsed) {\n return {\n nextValue: null,\n changed: true,\n migrated: false,\n clearedInvalid: true,\n }\n }\n\n const nextProvider = migrateProviderId(parsed.provider)\n const nextKey = composeModelKey(nextProvider, parsed.modelId)\n return {\n nextValue: nextKey || null,\n changed: (nextKey || null) !== (raw || null),\n migrated: parsed.provider !== nextProvider,\n clearedInvalid: false,\n }\n}\n\nfunction migrateCapabilitySelections(\n raw: string | null,\n): {\n ok: boolean\n nextRaw: string | null\n changed: boolean\n migratedKeys: number\n} {\n const parsed = parseJsonRecord(raw)\n if (!parsed.ok) {\n return {\n ok: false,\n nextRaw: null,\n changed: raw !== null,\n migratedKeys: 0,\n }\n }\n\n const deduped: CapabilitySelections = {}\n const priorities = new Map()\n let migratedKeys = 0\n\n for (const [modelKey, rawSelection] of Object.entries(parsed.value)) {\n if (!isRecord(rawSelection)) {\n return {\n ok: false,\n nextRaw: null,\n changed: raw !== null,\n migratedKeys: 0,\n }\n }\n\n const parsedKey = parseModelKeyStrict(modelKey)\n if (!parsedKey) {\n return {\n ok: false,\n nextRaw: null,\n changed: raw !== null,\n migratedKeys: 0,\n }\n }\n\n const nextKey = migrateModelKey(parsedKey.modelKey)\n if (nextKey !== parsedKey.modelKey) migratedKeys += 1\n\n const nextSelection: Record = {}\n for (const [field, value] of Object.entries(rawSelection)) {\n if (typeof value !== 'string' && typeof value !== 'number' && typeof value !== 'boolean') {\n return {\n ok: false,\n nextRaw: null,\n changed: raw !== null,\n migratedKeys: 0,\n }\n }\n nextSelection[field] = value\n }\n\n const sourcePriority = providerPriorityByOriginalKey(parsedKey.provider)\n const existingPriority = priorities.get(nextKey)\n if (existingPriority === undefined || sourcePriority > existingPriority) {\n deduped[nextKey] = nextSelection\n priorities.set(nextKey, sourcePriority)\n }\n }\n\n const nextRaw = Object.keys(deduped).length > 0 ? JSON.stringify(deduped) : null\n return {\n ok: true,\n nextRaw,\n changed: (raw || null) !== (nextRaw || null),\n migratedKeys,\n }\n}\n\nfunction toIndexNumber(value: number | string): number {\n if (typeof value === 'number') return value\n return Number.parseInt(value, 10)\n}\n\nfunction hasRequiredGraphArtifactUniqueIndex(rows: MysqlIndexRow[]): boolean {\n const indexColumns = new Map>()\n for (const row of rows) {\n const seq = toIndexNumber(row.Seq_in_index)\n const nonUnique = toIndexNumber(row.Non_unique)\n if (!Number.isFinite(seq) || !Number.isFinite(nonUnique)) continue\n const key = row.Key_name\n const list = indexColumns.get(key) || []\n list.push({\n seq,\n column: row.Column_name,\n nonUnique,\n })\n indexColumns.set(key, list)\n }\n\n for (const entries of indexColumns.values()) {\n if (entries.length !== REQUIRED_GRAPH_ARTIFACT_UNIQUE_COLUMNS.length) continue\n const sorted = entries.sort((a, b) => a.seq - b.seq)\n if (sorted[0]?.nonUnique !== 0) continue\n const columns = sorted.map((entry) => entry.column)\n const match = columns.every((column, index) => column === REQUIRED_GRAPH_ARTIFACT_UNIQUE_COLUMNS[index])\n if (match) return true\n }\n return false\n}\n\nfunction toNumber(value: bigint | number): number {\n if (typeof value === 'bigint') return Number(value)\n return value\n}\n\nasync function loadGraphArtifactIndexes(): Promise {\n return await prisma.$queryRawUnsafe('SHOW INDEX FROM graph_artifacts')\n}\n\nasync function countGraphArtifactDuplicateGroups(): Promise {\n const rows = await prisma.$queryRawUnsafe(\n `SELECT COUNT(*) AS c\n FROM (\n SELECT 1\n FROM graph_artifacts\n WHERE stepKey IS NOT NULL\n GROUP BY runId, stepKey, artifactType, refId\n HAVING COUNT(*) > 1\n ) duplicate_groups`,\n )\n return rows.length > 0 ? toNumber(rows[0].c) : 0\n}\n\nasync function sampleGraphArtifactDuplicateGroups(limit: number): Promise {\n return await prisma.$queryRawUnsafe(\n `SELECT runId, stepKey, artifactType, refId, COUNT(*) AS c\n FROM graph_artifacts\n WHERE stepKey IS NOT NULL\n GROUP BY runId, stepKey, artifactType, refId\n HAVING c > 1\n LIMIT ${limit}`,\n )\n}\n\nasync function dedupeGraphArtifacts(): Promise {\n return await prisma.$executeRawUnsafe(\n `DELETE ga1 FROM graph_artifacts ga1\n JOIN graph_artifacts ga2\n ON ga1.runId = ga2.runId\n AND ga1.stepKey = ga2.stepKey\n AND ga1.artifactType = ga2.artifactType\n AND ga1.refId = ga2.refId\n AND (\n ga1.createdAt < ga2.createdAt\n OR (ga1.createdAt = ga2.createdAt AND ga1.id < ga2.id)\n )\n WHERE ga1.stepKey IS NOT NULL`,\n )\n}\n\nasync function addGraphArtifactUniqueIndex(): Promise {\n await prisma.$executeRawUnsafe(\n 'ALTER TABLE graph_artifacts ADD UNIQUE INDEX graph_artifacts_runId_stepKey_artifactType_refId_key (runId, stepKey, artifactType, refId)',\n )\n}\n\nasync function migrateUserPreferences(summary: MigrationSummary): Promise {\n const rows = await prisma.userPreference.findMany({\n select: {\n id: true,\n userId: true,\n customProviders: true,\n customModels: true,\n analysisModel: true,\n characterModel: true,\n locationModel: true,\n storyboardModel: true,\n editModel: true,\n videoModel: true,\n audioModel: true,\n lipSyncModel: true,\n capabilityDefaults: true,\n },\n }) as UserPreferenceRow[]\n\n summary.userPreference.scanned = rows.length\n\n for (const row of rows) {\n const updateData: UserPreferenceUpdateData = {}\n let changed = false\n\n const providerResult = migrateProviders(row.customProviders)\n if (!providerResult.ok) {\n updateData.customProviders = null\n changed = changed || row.customProviders !== null\n summary.userPreference.dirtyClearedProviders += 1\n } else if (providerResult.changed) {\n updateData.customProviders = providerResult.nextRaw\n changed = true\n summary.userPreference.migratedProviders += providerResult.migratedProviders\n summary.userPreference.providerCollisionsResolvedByBailian += providerResult.collisionsResolvedByBailian\n }\n\n const modelResult = migrateModels(row.customModels)\n if (!modelResult.ok) {\n updateData.customModels = null\n changed = changed || row.customModels !== null\n summary.userPreference.dirtyClearedModels += 1\n } else if (modelResult.changed) {\n updateData.customModels = modelResult.nextRaw\n changed = true\n summary.userPreference.migratedModels += modelResult.migratedModels\n summary.userPreference.modelCollisionsResolvedByBailian += modelResult.collisionsResolvedByBailian\n }\n\n const capabilityResult = migrateCapabilitySelections(row.capabilityDefaults)\n if (!capabilityResult.ok) {\n updateData.capabilityDefaults = null\n changed = changed || row.capabilityDefaults !== null\n summary.userPreference.dirtyClearedCapabilityDefaults += 1\n } else if (capabilityResult.changed) {\n updateData.capabilityDefaults = capabilityResult.nextRaw\n changed = true\n summary.userPreference.migratedCapabilityDefaultKeys += capabilityResult.migratedKeys\n }\n\n for (const field of DEFAULT_MODEL_FIELDS) {\n const fieldResult = migrateModelField(row[field])\n if (!fieldResult.changed) continue\n updateData[field] = fieldResult.nextValue\n changed = true\n if (fieldResult.migrated) {\n summary.userPreference.migratedDefaultModelFields += 1\n }\n if (fieldResult.clearedInvalid) {\n summary.userPreference.invalidModelFieldsCleared += 1\n }\n }\n\n if (!changed) continue\n summary.userPreference.updated += 1\n\n if (APPLY) {\n await prisma.userPreference.update({\n where: { id: row.id },\n data: updateData,\n })\n }\n }\n}\n\nasync function migrateNovelProjects(summary: MigrationSummary): Promise {\n const rows = await prisma.novelPromotionProject.findMany({\n select: {\n id: true,\n projectId: true,\n analysisModel: true,\n characterModel: true,\n locationModel: true,\n storyboardModel: true,\n editModel: true,\n videoModel: true,\n capabilityOverrides: true,\n },\n }) as NovelProjectRow[]\n\n summary.novelPromotionProject.scanned = rows.length\n\n for (const row of rows) {\n const updateData: NovelProjectUpdateData = {}\n let changed = false\n\n for (const field of PROJECT_MODEL_FIELDS) {\n const fieldResult = migrateModelField(row[field])\n if (!fieldResult.changed) continue\n updateData[field] = fieldResult.nextValue\n changed = true\n if (fieldResult.migrated) {\n summary.novelPromotionProject.migratedModelFields += 1\n }\n if (fieldResult.clearedInvalid) {\n summary.novelPromotionProject.invalidModelFieldsCleared += 1\n }\n }\n\n const capabilityResult = migrateCapabilitySelections(row.capabilityOverrides)\n if (!capabilityResult.ok) {\n updateData.capabilityOverrides = null\n changed = changed || row.capabilityOverrides !== null\n summary.novelPromotionProject.dirtyClearedCapabilityOverrides += 1\n } else if (capabilityResult.changed) {\n updateData.capabilityOverrides = capabilityResult.nextRaw\n changed = true\n summary.novelPromotionProject.migratedCapabilityOverrideKeys += capabilityResult.migratedKeys\n }\n\n if (!changed) continue\n summary.novelPromotionProject.updated += 1\n\n if (APPLY) {\n await prisma.novelPromotionProject.update({\n where: { id: row.id },\n data: updateData,\n })\n }\n }\n}\n\nasync function migrateGraphArtifacts(summary: MigrationSummary): Promise {\n const beforeIndexes = await loadGraphArtifactIndexes()\n const hasRequiredBefore = hasRequiredGraphArtifactUniqueIndex(beforeIndexes)\n const duplicateGroupsBefore = await countGraphArtifactDuplicateGroups()\n const duplicateGroupSamples = await sampleGraphArtifactDuplicateGroups(20)\n\n summary.graphArtifacts.hasRequiredUniqueIndexBefore = hasRequiredBefore\n summary.graphArtifacts.duplicateGroupsBefore = duplicateGroupsBefore\n summary.graphArtifacts.duplicateGroupSamples = duplicateGroupSamples.map((row) => ({\n runId: row.runId,\n stepKey: row.stepKey,\n artifactType: row.artifactType,\n refId: row.refId,\n count: toNumber(row.c),\n }))\n\n if (APPLY && duplicateGroupsBefore > 0) {\n const deleted = await dedupeGraphArtifacts()\n summary.graphArtifacts.deletedRowsForDedup = deleted\n }\n\n const duplicateGroupsAfter = APPLY ? await countGraphArtifactDuplicateGroups() : duplicateGroupsBefore\n summary.graphArtifacts.duplicateGroupsAfter = duplicateGroupsAfter\n\n if (APPLY && !hasRequiredBefore) {\n if (duplicateGroupsAfter > 0) {\n throw new Error(\n `GRAPH_ARTIFACT_DEDUPE_INCOMPLETE: still has ${duplicateGroupsAfter} duplicate groups, unique index not added`,\n )\n }\n await addGraphArtifactUniqueIndex()\n summary.graphArtifacts.indexAdded = true\n }\n\n const afterIndexes = await loadGraphArtifactIndexes()\n summary.graphArtifacts.hasRequiredUniqueIndexAfter = hasRequiredGraphArtifactUniqueIndex(afterIndexes)\n if (APPLY && !summary.graphArtifacts.hasRequiredUniqueIndexAfter) {\n throw new Error('GRAPH_ARTIFACT_UNIQUE_INDEX_MISSING_AFTER_MIGRATION')\n }\n}\n\nasync function main() {\n const summary: MigrationSummary = {\n mode: MODE,\n userPreference: {\n scanned: 0,\n updated: 0,\n dirtyClearedProviders: 0,\n dirtyClearedModels: 0,\n dirtyClearedCapabilityDefaults: 0,\n migratedProviders: 0,\n migratedModels: 0,\n migratedDefaultModelFields: 0,\n migratedCapabilityDefaultKeys: 0,\n modelCollisionsResolvedByBailian: 0,\n providerCollisionsResolvedByBailian: 0,\n invalidModelFieldsCleared: 0,\n },\n novelPromotionProject: {\n scanned: 0,\n updated: 0,\n migratedModelFields: 0,\n migratedCapabilityOverrideKeys: 0,\n invalidModelFieldsCleared: 0,\n dirtyClearedCapabilityOverrides: 0,\n },\n graphArtifacts: {\n hasRequiredUniqueIndexBefore: false,\n duplicateGroupsBefore: 0,\n duplicateGroupSamples: [],\n deletedRowsForDedup: 0,\n duplicateGroupsAfter: 0,\n indexAdded: false,\n hasRequiredUniqueIndexAfter: false,\n },\n }\n\n await migrateUserPreferences(summary)\n await migrateNovelProjects(summary)\n await migrateGraphArtifacts(summary)\n\n console.log(JSON.stringify(summary, null, 2))\n}\n\nmain()\n .then(async () => {\n await prisma.$disconnect()\n })\n .catch(async (error: unknown) => {\n console.error('[migrate-release-blockers] failed', error)\n await prisma.$disconnect()\n process.exit(1)\n })\n"
},
{
"path": "scripts/migrations/reports/model-config-migration-report.apply.json",
"content": "{\n \"generatedAt\": \"2026-02-12T12:51:54.103Z\",\n \"mode\": \"apply\",\n \"userPreference\": {\n \"scanned\": 7,\n \"updated\": 1,\n \"updatedCustomModels\": 1,\n \"updatedDefaultFields\": 2\n },\n \"novelPromotionProject\": {\n \"scanned\": 70,\n \"updated\": 65,\n \"updatedFields\": 189\n },\n \"issues\": [\n {\n \"table\": \"userPreference\",\n \"rowId\": \"0c3ddcbd-d6f7-4b57-90d0-3c2fe0d48c5c\",\n \"userId\": \"edb99912-94ce-403d-97bb-2a29d5d20cb6\",\n \"field\": \"analysisModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"anthropic/claude-sonnet-4.5\",\n \"message\": \"analysisModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"0c3ddcbd-d6f7-4b57-90d0-3c2fe0d48c5c\",\n \"userId\": \"edb99912-94ce-403d-97bb-2a29d5d20cb6\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"0c3ddcbd-d6f7-4b57-90d0-3c2fe0d48c5c\",\n \"userId\": \"edb99912-94ce-403d-97bb-2a29d5d20cb6\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"0c3ddcbd-d6f7-4b57-90d0-3c2fe0d48c5c\",\n \"userId\": \"edb99912-94ce-403d-97bb-2a29d5d20cb6\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"gemini-3-pro-image-preview\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"0c3ddcbd-d6f7-4b57-90d0-3c2fe0d48c5c\",\n \"userId\": \"edb99912-94ce-403d-97bb-2a29d5d20cb6\",\n \"field\": \"editModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"editModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"0c3ddcbd-d6f7-4b57-90d0-3c2fe0d48c5c\",\n \"userId\": \"edb99912-94ce-403d-97bb-2a29d5d20cb6\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-fast-251015\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"4c73593e-148d-4707-a2c2-087d2fd43a68\",\n \"userId\": \"bea0c5b7-73f0-4048-8cb9-e6ef18650a0f\",\n \"field\": \"analysisModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"google/gemini-3-pro-preview\",\n \"message\": \"analysisModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"4c73593e-148d-4707-a2c2-087d2fd43a68\",\n \"userId\": \"bea0c5b7-73f0-4048-8cb9-e6ef18650a0f\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"4c73593e-148d-4707-a2c2-087d2fd43a68\",\n \"userId\": \"bea0c5b7-73f0-4048-8cb9-e6ef18650a0f\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"4c73593e-148d-4707-a2c2-087d2fd43a68\",\n \"userId\": \"bea0c5b7-73f0-4048-8cb9-e6ef18650a0f\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"4c73593e-148d-4707-a2c2-087d2fd43a68\",\n \"userId\": \"bea0c5b7-73f0-4048-8cb9-e6ef18650a0f\",\n \"field\": \"editModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"editModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"4c73593e-148d-4707-a2c2-087d2fd43a68\",\n \"userId\": \"bea0c5b7-73f0-4048-8cb9-e6ef18650a0f\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-250528\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"b33bc1c7-0ba8-475e-87e0-ba6ef897c2d2\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"analysisModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"google/gemini-3-pro-preview\",\n \"message\": \"analysisModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"b33bc1c7-0ba8-475e-87e0-ba6ef897c2d2\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"b33bc1c7-0ba8-475e-87e0-ba6ef897c2d2\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"b33bc1c7-0ba8-475e-87e0-ba6ef897c2d2\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"b33bc1c7-0ba8-475e-87e0-ba6ef897c2d2\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"editModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"editModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"b33bc1c7-0ba8-475e-87e0-ba6ef897c2d2\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-250528\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"bb2a2aa3-29f9-4f85-9e93-6363faf17dc4\",\n \"userId\": \"5a28aff0-e370-4337-acf6-0f8ce86f346a\",\n \"field\": \"analysisModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"google/gemini-3-pro-preview\",\n \"message\": \"analysisModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"bb2a2aa3-29f9-4f85-9e93-6363faf17dc4\",\n \"userId\": \"5a28aff0-e370-4337-acf6-0f8ce86f346a\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"bb2a2aa3-29f9-4f85-9e93-6363faf17dc4\",\n \"userId\": \"5a28aff0-e370-4337-acf6-0f8ce86f346a\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"bb2a2aa3-29f9-4f85-9e93-6363faf17dc4\",\n \"userId\": \"5a28aff0-e370-4337-acf6-0f8ce86f346a\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"bb2a2aa3-29f9-4f85-9e93-6363faf17dc4\",\n \"userId\": \"5a28aff0-e370-4337-acf6-0f8ce86f346a\",\n \"field\": \"editModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"editModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"bb2a2aa3-29f9-4f85-9e93-6363faf17dc4\",\n \"userId\": \"5a28aff0-e370-4337-acf6-0f8ce86f346a\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-250528\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"04ebb651-dc90-4c69-99d4-343dd037a9b9\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"analysisModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"anthropic/claude-sonnet-4.5\",\n \"message\": \"analysisModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"04ebb651-dc90-4c69-99d4-343dd037a9b9\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"04ebb651-dc90-4c69-99d4-343dd037a9b9\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"04ebb651-dc90-4c69-99d4-343dd037a9b9\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"04ebb651-dc90-4c69-99d4-343dd037a9b9\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-250528\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"0574a234-e128-4196-a103-1fa8305f3ca0\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"0574a234-e128-4196-a103-1fa8305f3ca0\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-250528\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"0a235755-3a90-44b1-813d-d00d163662b9\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"0a235755-3a90-44b1-813d-d00d163662b9\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"0a235755-3a90-44b1-813d-d00d163662b9\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"gemini-3-pro-image-preview-batch\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"155e9723-b2ef-414f-8c5e-f9e1e34de536\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-fast-251015-batch\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"1979ed4c-0d16-4696-a9ea-4a15e0c4cd0f\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"1979ed4c-0d16-4696-a9ea-4a15e0c4cd0f\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"1a296d18-c305-4639-9b11-3c694cd0718d\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"1a296d18-c305-4639-9b11-3c694cd0718d\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"2f898804-cda1-417a-9f58-6d19fa0f134d\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"gemini-3.0-pro-image-portrait\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"374c975a-8223-453e-81a4-d110c7e4b685\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"374c975a-8223-453e-81a4-d110c7e4b685\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"374c975a-8223-453e-81a4-d110c7e4b685\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"gemini-3-pro-image-preview-batch\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"49909329-a771-4b18-a305-ff7d284bcf4e\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"49909329-a771-4b18-a305-ff7d284bcf4e\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"49909329-a771-4b18-a305-ff7d284bcf4e\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"49909329-a771-4b18-a305-ff7d284bcf4e\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-fast-251015-batch\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"4f0eaccc-5fe4-4440-bcc0-be70892626c5\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"4f0eaccc-5fe4-4440-bcc0-be70892626c5\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"55e56c35-f59f-4d32-9e99-d67b46846870\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"55e56c35-f59f-4d32-9e99-d67b46846870\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"5cdef710-6aef-445a-b217-2cdb6987a68a\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"5cdef710-6aef-445a-b217-2cdb6987a68a\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"5cdef710-6aef-445a-b217-2cdb6987a68a\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"5cdef710-6aef-445a-b217-2cdb6987a68a\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-fast-251015-batch\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"5d97db15-8e16-45dc-8fa7-5efa5a237a99\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"5d97db15-8e16-45dc-8fa7-5efa5a237a99\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"673acab3-9d09-4b1f-87cc-9cafbe60af80\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"673acab3-9d09-4b1f-87cc-9cafbe60af80\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"6bbb564a-14bd-4bc2-83e7-1a8e8612eb8e\",\n \"userId\": \"edb99912-94ce-403d-97bb-2a29d5d20cb6\",\n \"field\": \"analysisModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"anthropic/claude-sonnet-4.5\",\n \"message\": \"analysisModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"6bbb564a-14bd-4bc2-83e7-1a8e8612eb8e\",\n \"userId\": \"edb99912-94ce-403d-97bb-2a29d5d20cb6\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"6bbb564a-14bd-4bc2-83e7-1a8e8612eb8e\",\n \"userId\": \"edb99912-94ce-403d-97bb-2a29d5d20cb6\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"6bbb564a-14bd-4bc2-83e7-1a8e8612eb8e\",\n \"userId\": \"edb99912-94ce-403d-97bb-2a29d5d20cb6\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"gemini-3-pro-image-preview\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"6bbb564a-14bd-4bc2-83e7-1a8e8612eb8e\",\n \"userId\": \"edb99912-94ce-403d-97bb-2a29d5d20cb6\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-fast-251015\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"77b44773-1264-419f-b145-4c2a9cbbd977\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"77b44773-1264-419f-b145-4c2a9cbbd977\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"77b44773-1264-419f-b145-4c2a9cbbd977\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"77b44773-1264-419f-b145-4c2a9cbbd977\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-fast-251015-batch\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"788c5fc0-e3b6-48dd-a2e5-719069726a08\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"788c5fc0-e3b6-48dd-a2e5-719069726a08\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"gemini-3-pro-image-preview-batch\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"811a2cde-2a8e-4934-947b-18bf9d6331e0\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana-2k\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"88dbc942-e549-48f8-808b-53e57fbdb97f\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"88dbc942-e549-48f8-808b-53e57fbdb97f\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"8ef54b9c-0142-49ad-a9cd-ec94c5132731\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"8ef54b9c-0142-49ad-a9cd-ec94c5132731\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"8ef54b9c-0142-49ad-a9cd-ec94c5132731\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-fast-251015-batch\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"9b48f96d-7d46-4264-abfd-82ee338eecd7\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"b1013460-0798-43f2-87bd-7b018cca1d58\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana-2k\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"b8740c4d-0694-4106-9831-db665225a0c5\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"gemini-3-pro-image-preview-batch\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"b8740c4d-0694-4106-9831-db665225a0c5\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"b8740c4d-0694-4106-9831-db665225a0c5\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"gemini-3-pro-image-preview-batch\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"b9f3f933-13d6-461a-bb80-1e94b48c855d\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"b9f3f933-13d6-461a-bb80-1e94b48c855d\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"b9f3f933-13d6-461a-bb80-1e94b48c855d\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"gemini-3-pro-image-preview-batch\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"b9f3f933-13d6-461a-bb80-1e94b48c855d\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-fast-251015-batch\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"bc31aa21-9eec-47d2-9853-e4b34d926b6a\",\n \"userId\": \"5a28aff0-e370-4337-acf6-0f8ce86f346a\",\n \"field\": \"analysisModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"google/gemini-3-pro-preview\",\n \"message\": \"analysisModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"bc31aa21-9eec-47d2-9853-e4b34d926b6a\",\n \"userId\": \"5a28aff0-e370-4337-acf6-0f8ce86f346a\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"bc31aa21-9eec-47d2-9853-e4b34d926b6a\",\n \"userId\": \"5a28aff0-e370-4337-acf6-0f8ce86f346a\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"bc31aa21-9eec-47d2-9853-e4b34d926b6a\",\n \"userId\": \"5a28aff0-e370-4337-acf6-0f8ce86f346a\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"bc31aa21-9eec-47d2-9853-e4b34d926b6a\",\n \"userId\": \"5a28aff0-e370-4337-acf6-0f8ce86f346a\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-250528\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"bcf736cd-aeb9-49ce-bb36-d878400bbfd6\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"bcf736cd-aeb9-49ce-bb36-d878400bbfd6\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"bcf736cd-aeb9-49ce-bb36-d878400bbfd6\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"bcf736cd-aeb9-49ce-bb36-d878400bbfd6\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-fast-251015-batch\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"c38b1c73-9baa-4a03-86b7-5f6dbecc0e1e\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"c38b1c73-9baa-4a03-86b7-5f6dbecc0e1e\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"c61f1a58-42e8-4989-968b-5a373ae77e1a\",\n \"userId\": \"bea0c5b7-73f0-4048-8cb9-e6ef18650a0f\",\n \"field\": \"analysisModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"google/gemini-3-pro-preview\",\n \"message\": \"analysisModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"c61f1a58-42e8-4989-968b-5a373ae77e1a\",\n \"userId\": \"bea0c5b7-73f0-4048-8cb9-e6ef18650a0f\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"c61f1a58-42e8-4989-968b-5a373ae77e1a\",\n \"userId\": \"bea0c5b7-73f0-4048-8cb9-e6ef18650a0f\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"c61f1a58-42e8-4989-968b-5a373ae77e1a\",\n \"userId\": \"bea0c5b7-73f0-4048-8cb9-e6ef18650a0f\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"c61f1a58-42e8-4989-968b-5a373ae77e1a\",\n \"userId\": \"bea0c5b7-73f0-4048-8cb9-e6ef18650a0f\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-250528\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"ca374b6b-5666-4940-a859-23761fbacd62\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"analysisModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"openai/gpt-5.2\",\n \"message\": \"analysisModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"d1bbeef0-94d5-4247-b90f-105b432649e1\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"d1bbeef0-94d5-4247-b90f-105b432649e1\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"d2c7979d-839b-4baa-bbf0-c726f2cc09bb\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"imagen-4.0-ultra-generate-001\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"d2c7979d-839b-4baa-bbf0-c726f2cc09bb\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"d2c7979d-839b-4baa-bbf0-c726f2cc09bb\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"gemini-3-pro-image-preview-batch\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"dc0669ec-effc-4626-a038-bbf48994d65a\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"dc0669ec-effc-4626-a038-bbf48994d65a\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"gemini-3-pro-image-preview-batch\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"dc0669ec-effc-4626-a038-bbf48994d65a\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-fast-251015-batch\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"de428a7c-e719-4cda-874a-7c54878a14bc\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana-2k\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"ec2526c0-1acd-4081-b13d-019d5d425710\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"ec2526c0-1acd-4081-b13d-019d5d425710\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"ec2526c0-1acd-4081-b13d-019d5d425710\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"gemini-3-pro-image-preview-batch\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"ed2d93ad-5cfa-4bd0-b373-a65aae679ef5\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"ee5e7440-c825-4c0c-b903-0a51d983db6f\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"ee5e7440-c825-4c0c-b903-0a51d983db6f\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"f36d983d-ba52-4849-bcda-0d710b2e04f9\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"nano-banana-pro\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"f36d983d-ba52-4849-bcda-0d710b2e04f9\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"editModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"nano-banana-pro\",\n \"message\": \"editModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"f38e00d3-09b0-47a5-8caf-71d491bae4d1\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"analysisModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"google/gemini-3-pro-preview\",\n \"message\": \"analysisModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"f38e00d3-09b0-47a5-8caf-71d491bae4d1\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"f38e00d3-09b0-47a5-8caf-71d491bae4d1\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"f38e00d3-09b0-47a5-8caf-71d491bae4d1\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"f38e00d3-09b0-47a5-8caf-71d491bae4d1\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-250528\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"f45efd24-37b1-4ca6-8808-0b5c9de205f1\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"gemini-3-pro-image-preview-batch\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"f45efd24-37b1-4ca6-8808-0b5c9de205f1\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"f45efd24-37b1-4ca6-8808-0b5c9de205f1\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"gemini-3-pro-image-preview-batch\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"ff9129c7-f76f-4ab0-8fae-84be89b62390\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"ff9129c7-f76f-4ab0-8fae-84be89b62390\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"ff9129c7-f76f-4ab0-8fae-84be89b62390\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"gemini-3-pro-image-preview-batch\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n }\n ]\n}\n"
},
{
"path": "scripts/migrations/reports/model-config-migration-report.post-alias-apply.json",
"content": "{\n \"generatedAt\": \"2026-02-12T12:53:18.381Z\",\n \"mode\": \"apply\",\n \"userPreference\": {\n \"scanned\": 7,\n \"updated\": 4,\n \"updatedCustomModels\": 0,\n \"updatedDefaultFields\": 24\n },\n \"novelPromotionProject\": {\n \"scanned\": 70,\n \"updated\": 40,\n \"updatedFields\": 106\n },\n \"issues\": []\n}\n"
},
{
"path": "scripts/migrations/reports/model-config-migration-report.post-alias-dryrun.json",
"content": "{\n \"generatedAt\": \"2026-02-12T12:53:12.288Z\",\n \"mode\": \"dry-run\",\n \"userPreference\": {\n \"scanned\": 7,\n \"updated\": 4,\n \"updatedCustomModels\": 0,\n \"updatedDefaultFields\": 24\n },\n \"novelPromotionProject\": {\n \"scanned\": 70,\n \"updated\": 40,\n \"updatedFields\": 106\n },\n \"issues\": []\n}\n"
},
{
"path": "scripts/migrations/reports/model-config-migration-report.pre-apply.json",
"content": "{\n \"generatedAt\": \"2026-02-12T12:51:46.934Z\",\n \"mode\": \"dry-run\",\n \"userPreference\": {\n \"scanned\": 7,\n \"updated\": 1,\n \"updatedCustomModels\": 1,\n \"updatedDefaultFields\": 2\n },\n \"novelPromotionProject\": {\n \"scanned\": 70,\n \"updated\": 65,\n \"updatedFields\": 189\n },\n \"issues\": [\n {\n \"table\": \"userPreference\",\n \"rowId\": \"0c3ddcbd-d6f7-4b57-90d0-3c2fe0d48c5c\",\n \"userId\": \"edb99912-94ce-403d-97bb-2a29d5d20cb6\",\n \"field\": \"analysisModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"anthropic/claude-sonnet-4.5\",\n \"message\": \"analysisModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"0c3ddcbd-d6f7-4b57-90d0-3c2fe0d48c5c\",\n \"userId\": \"edb99912-94ce-403d-97bb-2a29d5d20cb6\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"0c3ddcbd-d6f7-4b57-90d0-3c2fe0d48c5c\",\n \"userId\": \"edb99912-94ce-403d-97bb-2a29d5d20cb6\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"0c3ddcbd-d6f7-4b57-90d0-3c2fe0d48c5c\",\n \"userId\": \"edb99912-94ce-403d-97bb-2a29d5d20cb6\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"gemini-3-pro-image-preview\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"0c3ddcbd-d6f7-4b57-90d0-3c2fe0d48c5c\",\n \"userId\": \"edb99912-94ce-403d-97bb-2a29d5d20cb6\",\n \"field\": \"editModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"editModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"0c3ddcbd-d6f7-4b57-90d0-3c2fe0d48c5c\",\n \"userId\": \"edb99912-94ce-403d-97bb-2a29d5d20cb6\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-fast-251015\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"4c73593e-148d-4707-a2c2-087d2fd43a68\",\n \"userId\": \"bea0c5b7-73f0-4048-8cb9-e6ef18650a0f\",\n \"field\": \"analysisModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"google/gemini-3-pro-preview\",\n \"message\": \"analysisModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"4c73593e-148d-4707-a2c2-087d2fd43a68\",\n \"userId\": \"bea0c5b7-73f0-4048-8cb9-e6ef18650a0f\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"4c73593e-148d-4707-a2c2-087d2fd43a68\",\n \"userId\": \"bea0c5b7-73f0-4048-8cb9-e6ef18650a0f\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"4c73593e-148d-4707-a2c2-087d2fd43a68\",\n \"userId\": \"bea0c5b7-73f0-4048-8cb9-e6ef18650a0f\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"4c73593e-148d-4707-a2c2-087d2fd43a68\",\n \"userId\": \"bea0c5b7-73f0-4048-8cb9-e6ef18650a0f\",\n \"field\": \"editModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"editModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"4c73593e-148d-4707-a2c2-087d2fd43a68\",\n \"userId\": \"bea0c5b7-73f0-4048-8cb9-e6ef18650a0f\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-250528\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"b33bc1c7-0ba8-475e-87e0-ba6ef897c2d2\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"analysisModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"google/gemini-3-pro-preview\",\n \"message\": \"analysisModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"b33bc1c7-0ba8-475e-87e0-ba6ef897c2d2\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"b33bc1c7-0ba8-475e-87e0-ba6ef897c2d2\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"b33bc1c7-0ba8-475e-87e0-ba6ef897c2d2\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"b33bc1c7-0ba8-475e-87e0-ba6ef897c2d2\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"editModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"editModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"b33bc1c7-0ba8-475e-87e0-ba6ef897c2d2\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-250528\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"bb2a2aa3-29f9-4f85-9e93-6363faf17dc4\",\n \"userId\": \"5a28aff0-e370-4337-acf6-0f8ce86f346a\",\n \"field\": \"analysisModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"google/gemini-3-pro-preview\",\n \"message\": \"analysisModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"bb2a2aa3-29f9-4f85-9e93-6363faf17dc4\",\n \"userId\": \"5a28aff0-e370-4337-acf6-0f8ce86f346a\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"bb2a2aa3-29f9-4f85-9e93-6363faf17dc4\",\n \"userId\": \"5a28aff0-e370-4337-acf6-0f8ce86f346a\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"bb2a2aa3-29f9-4f85-9e93-6363faf17dc4\",\n \"userId\": \"5a28aff0-e370-4337-acf6-0f8ce86f346a\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"bb2a2aa3-29f9-4f85-9e93-6363faf17dc4\",\n \"userId\": \"5a28aff0-e370-4337-acf6-0f8ce86f346a\",\n \"field\": \"editModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"editModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"userPreference\",\n \"rowId\": \"bb2a2aa3-29f9-4f85-9e93-6363faf17dc4\",\n \"userId\": \"5a28aff0-e370-4337-acf6-0f8ce86f346a\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-250528\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"04ebb651-dc90-4c69-99d4-343dd037a9b9\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"analysisModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"anthropic/claude-sonnet-4.5\",\n \"message\": \"analysisModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"04ebb651-dc90-4c69-99d4-343dd037a9b9\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"04ebb651-dc90-4c69-99d4-343dd037a9b9\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"04ebb651-dc90-4c69-99d4-343dd037a9b9\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"04ebb651-dc90-4c69-99d4-343dd037a9b9\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-250528\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"0574a234-e128-4196-a103-1fa8305f3ca0\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"0574a234-e128-4196-a103-1fa8305f3ca0\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-250528\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"0a235755-3a90-44b1-813d-d00d163662b9\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"0a235755-3a90-44b1-813d-d00d163662b9\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"0a235755-3a90-44b1-813d-d00d163662b9\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"gemini-3-pro-image-preview-batch\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"155e9723-b2ef-414f-8c5e-f9e1e34de536\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-fast-251015-batch\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"1979ed4c-0d16-4696-a9ea-4a15e0c4cd0f\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"1979ed4c-0d16-4696-a9ea-4a15e0c4cd0f\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"1a296d18-c305-4639-9b11-3c694cd0718d\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"1a296d18-c305-4639-9b11-3c694cd0718d\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"2f898804-cda1-417a-9f58-6d19fa0f134d\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"gemini-3.0-pro-image-portrait\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"374c975a-8223-453e-81a4-d110c7e4b685\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"374c975a-8223-453e-81a4-d110c7e4b685\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"374c975a-8223-453e-81a4-d110c7e4b685\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"gemini-3-pro-image-preview-batch\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"49909329-a771-4b18-a305-ff7d284bcf4e\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"49909329-a771-4b18-a305-ff7d284bcf4e\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"49909329-a771-4b18-a305-ff7d284bcf4e\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"49909329-a771-4b18-a305-ff7d284bcf4e\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-fast-251015-batch\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"4f0eaccc-5fe4-4440-bcc0-be70892626c5\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"4f0eaccc-5fe4-4440-bcc0-be70892626c5\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"55e56c35-f59f-4d32-9e99-d67b46846870\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"55e56c35-f59f-4d32-9e99-d67b46846870\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"5cdef710-6aef-445a-b217-2cdb6987a68a\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"5cdef710-6aef-445a-b217-2cdb6987a68a\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"5cdef710-6aef-445a-b217-2cdb6987a68a\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"5cdef710-6aef-445a-b217-2cdb6987a68a\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-fast-251015-batch\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"5d97db15-8e16-45dc-8fa7-5efa5a237a99\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"5d97db15-8e16-45dc-8fa7-5efa5a237a99\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"673acab3-9d09-4b1f-87cc-9cafbe60af80\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"673acab3-9d09-4b1f-87cc-9cafbe60af80\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"6bbb564a-14bd-4bc2-83e7-1a8e8612eb8e\",\n \"userId\": \"edb99912-94ce-403d-97bb-2a29d5d20cb6\",\n \"field\": \"analysisModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"anthropic/claude-sonnet-4.5\",\n \"message\": \"analysisModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"6bbb564a-14bd-4bc2-83e7-1a8e8612eb8e\",\n \"userId\": \"edb99912-94ce-403d-97bb-2a29d5d20cb6\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"6bbb564a-14bd-4bc2-83e7-1a8e8612eb8e\",\n \"userId\": \"edb99912-94ce-403d-97bb-2a29d5d20cb6\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"6bbb564a-14bd-4bc2-83e7-1a8e8612eb8e\",\n \"userId\": \"edb99912-94ce-403d-97bb-2a29d5d20cb6\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"gemini-3-pro-image-preview\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"6bbb564a-14bd-4bc2-83e7-1a8e8612eb8e\",\n \"userId\": \"edb99912-94ce-403d-97bb-2a29d5d20cb6\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-fast-251015\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"77b44773-1264-419f-b145-4c2a9cbbd977\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"77b44773-1264-419f-b145-4c2a9cbbd977\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"77b44773-1264-419f-b145-4c2a9cbbd977\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"77b44773-1264-419f-b145-4c2a9cbbd977\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-fast-251015-batch\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"788c5fc0-e3b6-48dd-a2e5-719069726a08\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"788c5fc0-e3b6-48dd-a2e5-719069726a08\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"gemini-3-pro-image-preview-batch\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"811a2cde-2a8e-4934-947b-18bf9d6331e0\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana-2k\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"88dbc942-e549-48f8-808b-53e57fbdb97f\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"88dbc942-e549-48f8-808b-53e57fbdb97f\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"8ef54b9c-0142-49ad-a9cd-ec94c5132731\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"8ef54b9c-0142-49ad-a9cd-ec94c5132731\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"8ef54b9c-0142-49ad-a9cd-ec94c5132731\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-fast-251015-batch\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"9b48f96d-7d46-4264-abfd-82ee338eecd7\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"b1013460-0798-43f2-87bd-7b018cca1d58\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana-2k\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"b8740c4d-0694-4106-9831-db665225a0c5\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"gemini-3-pro-image-preview-batch\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"b8740c4d-0694-4106-9831-db665225a0c5\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"b8740c4d-0694-4106-9831-db665225a0c5\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"gemini-3-pro-image-preview-batch\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"b9f3f933-13d6-461a-bb80-1e94b48c855d\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"b9f3f933-13d6-461a-bb80-1e94b48c855d\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"b9f3f933-13d6-461a-bb80-1e94b48c855d\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"gemini-3-pro-image-preview-batch\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"b9f3f933-13d6-461a-bb80-1e94b48c855d\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-fast-251015-batch\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"bc31aa21-9eec-47d2-9853-e4b34d926b6a\",\n \"userId\": \"5a28aff0-e370-4337-acf6-0f8ce86f346a\",\n \"field\": \"analysisModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"google/gemini-3-pro-preview\",\n \"message\": \"analysisModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"bc31aa21-9eec-47d2-9853-e4b34d926b6a\",\n \"userId\": \"5a28aff0-e370-4337-acf6-0f8ce86f346a\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"bc31aa21-9eec-47d2-9853-e4b34d926b6a\",\n \"userId\": \"5a28aff0-e370-4337-acf6-0f8ce86f346a\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"bc31aa21-9eec-47d2-9853-e4b34d926b6a\",\n \"userId\": \"5a28aff0-e370-4337-acf6-0f8ce86f346a\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"bc31aa21-9eec-47d2-9853-e4b34d926b6a\",\n \"userId\": \"5a28aff0-e370-4337-acf6-0f8ce86f346a\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-250528\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"bcf736cd-aeb9-49ce-bb36-d878400bbfd6\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"bcf736cd-aeb9-49ce-bb36-d878400bbfd6\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"bcf736cd-aeb9-49ce-bb36-d878400bbfd6\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"bcf736cd-aeb9-49ce-bb36-d878400bbfd6\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-fast-251015-batch\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"c38b1c73-9baa-4a03-86b7-5f6dbecc0e1e\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"c38b1c73-9baa-4a03-86b7-5f6dbecc0e1e\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"c61f1a58-42e8-4989-968b-5a373ae77e1a\",\n \"userId\": \"bea0c5b7-73f0-4048-8cb9-e6ef18650a0f\",\n \"field\": \"analysisModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"google/gemini-3-pro-preview\",\n \"message\": \"analysisModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"c61f1a58-42e8-4989-968b-5a373ae77e1a\",\n \"userId\": \"bea0c5b7-73f0-4048-8cb9-e6ef18650a0f\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"c61f1a58-42e8-4989-968b-5a373ae77e1a\",\n \"userId\": \"bea0c5b7-73f0-4048-8cb9-e6ef18650a0f\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"c61f1a58-42e8-4989-968b-5a373ae77e1a\",\n \"userId\": \"bea0c5b7-73f0-4048-8cb9-e6ef18650a0f\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"c61f1a58-42e8-4989-968b-5a373ae77e1a\",\n \"userId\": \"bea0c5b7-73f0-4048-8cb9-e6ef18650a0f\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-250528\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"ca374b6b-5666-4940-a859-23761fbacd62\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"analysisModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"openai/gpt-5.2\",\n \"message\": \"analysisModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"d1bbeef0-94d5-4247-b90f-105b432649e1\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"d1bbeef0-94d5-4247-b90f-105b432649e1\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"d2c7979d-839b-4baa-bbf0-c726f2cc09bb\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"imagen-4.0-ultra-generate-001\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"d2c7979d-839b-4baa-bbf0-c726f2cc09bb\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"d2c7979d-839b-4baa-bbf0-c726f2cc09bb\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"gemini-3-pro-image-preview-batch\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"dc0669ec-effc-4626-a038-bbf48994d65a\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"dc0669ec-effc-4626-a038-bbf48994d65a\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"gemini-3-pro-image-preview-batch\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"dc0669ec-effc-4626-a038-bbf48994d65a\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-fast-251015-batch\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"de428a7c-e719-4cda-874a-7c54878a14bc\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana-2k\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"ec2526c0-1acd-4081-b13d-019d5d425710\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"ec2526c0-1acd-4081-b13d-019d5d425710\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"ec2526c0-1acd-4081-b13d-019d5d425710\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"gemini-3-pro-image-preview-batch\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"ed2d93ad-5cfa-4bd0-b373-a65aae679ef5\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"ee5e7440-c825-4c0c-b903-0a51d983db6f\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"ee5e7440-c825-4c0c-b903-0a51d983db6f\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"f36d983d-ba52-4849-bcda-0d710b2e04f9\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"nano-banana-pro\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"f36d983d-ba52-4849-bcda-0d710b2e04f9\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"editModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"nano-banana-pro\",\n \"message\": \"editModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"f38e00d3-09b0-47a5-8caf-71d491bae4d1\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"analysisModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"google/gemini-3-pro-preview\",\n \"message\": \"analysisModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"f38e00d3-09b0-47a5-8caf-71d491bae4d1\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"f38e00d3-09b0-47a5-8caf-71d491bae4d1\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"f38e00d3-09b0-47a5-8caf-71d491bae4d1\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"banana\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"f38e00d3-09b0-47a5-8caf-71d491bae4d1\",\n \"userId\": \"91c9fe74-556d-4036-a2a0-ebd3336fd8d8\",\n \"field\": \"videoModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"doubao-seedance-1-0-pro-250528\",\n \"message\": \"videoModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"f45efd24-37b1-4ca6-8808-0b5c9de205f1\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"gemini-3-pro-image-preview-batch\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"f45efd24-37b1-4ca6-8808-0b5c9de205f1\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"f45efd24-37b1-4ca6-8808-0b5c9de205f1\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"gemini-3-pro-image-preview-batch\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"ff9129c7-f76f-4ab0-8fae-84be89b62390\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"characterModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"characterModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"ff9129c7-f76f-4ab0-8fae-84be89b62390\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"locationModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"seedream4.5\",\n \"message\": \"locationModel legacy modelId cannot be mapped\"\n },\n {\n \"table\": \"novelPromotionProject\",\n \"rowId\": \"ff9129c7-f76f-4ab0-8fae-84be89b62390\",\n \"userId\": \"3d84c341-87d7-4165-971d-a3f6c576aa21\",\n \"field\": \"storyboardModel\",\n \"kind\": \"LEGACY_MODEL_ID_NOT_FOUND\",\n \"rawValue\": \"gemini-3-pro-image-preview-batch\",\n \"message\": \"storyboardModel legacy modelId cannot be mapped\"\n }\n ]\n}\n"
},
{
"path": "scripts/task-error-stats.ts",
"content": "import { logInfo as _ulogInfo, logError as _ulogError } from '@/lib/logging/core'\nimport { prisma } from '@/lib/prisma'\n\nfunction parseMinutesArg() {\n const raw = process.argv.find((arg) => arg.startsWith('--minutes='))\n const value = raw ? Number.parseInt(raw.split('=')[1], 10) : 5\n return Number.isFinite(value) && value > 0 ? value : 5\n}\n\nasync function main() {\n const minutes = parseMinutesArg()\n const since = new Date(Date.now() - minutes * 60_000)\n\n const rows = await prisma.task.groupBy({\n by: ['errorCode'],\n where: {\n status: 'failed',\n finishedAt: { gte: since },\n },\n _count: {\n _all: true,\n },\n orderBy: {\n _count: {\n errorCode: 'desc',\n },\n },\n })\n\n const total = rows.reduce((sum: number, row) => sum + (row._count?._all || 0), 0)\n\n _ulogInfo(`[TaskErrorStats] window=${minutes}m failed_total=${total}`)\n if (!rows.length) {\n _ulogInfo('No failed tasks in the selected window.')\n return\n }\n\n for (const row of rows) {\n const code = row.errorCode || 'UNKNOWN'\n const count = row?._count?._all || 0\n const ratio = total > 0 ? ((count / total) * 100).toFixed(1) : '0.0'\n _ulogInfo(`${code}\\t${count}\\t${ratio}%`)\n }\n}\n\nmain()\n .catch((error) => {\n _ulogError('[TaskErrorStats] failed:', error?.message || error)\n process.exit(1)\n })\n .finally(async () => {\n await prisma.$disconnect()\n })\n"
},
{
"path": "scripts/test-full-image-flow.ts",
"content": "/**\n * 模拟完整的图像生成和显示流程\n * 运行: npx tsx scripts/test-full-image-flow.ts\n */\nimport { config } from 'dotenv'\nconfig()\n\nimport { uploadObject, getStorageProvider } from '../src/lib/storage'\nimport { extractStorageKeyFromLegacyValue, resolveMediaRefFromLegacyValue, getMediaObjectByPublicId } from '../src/lib/media/service'\nimport { attachMediaFieldsToProject } from '../src/lib/media/attach'\nimport { randomUUID } from 'crypto'\n\nasync function testFullImageFlow() {\n console.log('🧪 模拟完整图像生成和显示流程...\\n')\n\n const provider = getStorageProvider()\n console.log(`存储类型: ${provider.kind}\\n`)\n\n // 1. 模拟图像生成后的上传\n console.log('1️⃣ 模拟图像生成后上传:')\n const testKey = `images/location-${randomUUID()}.jpg`\n const testImageContent = Buffer.from('fake-generated-image-data')\n \n const storedKey = await uploadObject(testImageContent, testKey)\n console.log(` ✅ 上传成功，返回 key: ${storedKey}`)\n\n // 2. 模拟数据库存储（存储 key）\n console.log('\\n2️⃣ 模拟数据库存储:')\n const mockDbLocation = {\n id: 'loc-test-123',\n name: '测试场景',\n images: [\n {\n id: 'img-1',\n imageUrl: storedKey, // 存储的是 key，不是完整 URL\n imageIndex: 0,\n }\n ]\n }\n console.log(` 存储的 imageUrl: ${storedKey}`)\n\n // 3. 测试 extractStorageKeyFromLegacyValue\n console.log('\\n3️⃣ 测试 extractStorageKeyFromLegacyValue:')\n const extractedKey = extractStorageKeyFromLegacyValue(storedKey)\n console.log(` 输入: ${storedKey}`)\n console.log(` 输出: ${extractedKey}`)\n if (extractedKey) {\n console.log(` ✅ 成功提取 storageKey`)\n } else {\n console.log(` ❌ 未能提取 storageKey - 这是问题所在！`)\n }\n\n // 4. 测试 resolveMediaRefFromLegacyValue（创建 MediaObject）\n console.log('\\n4️⃣ 测试 resolveMediaRefFromLegacyValue:')\n try {\n const mediaRef = await resolveMediaRefFromLegacyValue(storedKey)\n if (mediaRef) {\n console.log(` ✅ MediaObject 创建/获取成功`)\n console.log(` id: ${mediaRef.id}`)\n console.log(` publicId: ${mediaRef.publicId}`)\n console.log(` url: ${mediaRef.url}`)\n console.log(` storageKey: ${mediaRef.storageKey}`)\n } else {\n console.log(` ❌ MediaRef 为 null`)\n }\n } catch (error) {\n console.log(` ❌ 失败:`, error)\n }\n\n // 5. 测试 attachMediaFieldsToProject（完整流程）\n console.log('\\n5️⃣ 测试 attachMediaFieldsToProject（API 层转换）:')\n try {\n const mockProject = {\n id: 'proj-test',\n locations: [mockDbLocation]\n }\n \n const result = await attachMediaFieldsToProject(mockProject)\n const location = result.locations?.[0]\n const image = location?.images?.[0]\n \n console.log(` 转换后的 imageUrl: ${image?.imageUrl}`)\n \n if (image?.imageUrl?.startsWith('/m/')) {\n console.log(` ✅ 正确生成了 /m/ 格式的 URL`)\n \n // 提取 publicId\n const publicId = image.imageUrl.replace('/m/', '').split('?')[0]\n console.log(` publicId: ${publicId}`)\n \n // 验证 MediaObject 存在\n const media = await getMediaObjectByPublicId(publicId)\n if (media) {\n console.log(` ✅ MediaObject 存在，storageKey: ${media.storageKey}`)\n } else {\n console.log(` ❌ MediaObject 不存在！`)\n }\n } else if (image?.imageUrl?.startsWith('http')) {\n console.log(` ⚠️ 返回了完整 HTTP URL: ${image.imageUrl}`)\n } else if (!image?.imageUrl) {\n console.log(` ❌ imageUrl 为空！`)\n } else {\n console.log(` ⚠️ URL 格式: ${image.imageUrl}`)\n }\n } catch (error) {\n console.log(` ❌ 失败:`, error)\n }\n\n // 6. 测试访问 /m/ URL\n console.log('\\n6️⃣ 测试访问 /m/ URL:')\n try {\n const mockProject = {\n id: 'proj-test',\n locations: [mockDbLocation]\n }\n \n const result = await attachMediaFieldsToProject(mockProject)\n const imageUrl = result.locations?.[0]?.images?.[0]?.imageUrl\n \n if (imageUrl?.startsWith('/m/')) {\n const fullUrl = `http://localhost:3000${imageUrl}`\n console.log(` 尝试访问: ${fullUrl}`)\n \n try {\n const response = await fetch(fullUrl, { redirect: 'manual' })\n console.log(` 状态: ${response.status}`)\n \n if (response.status === 200) {\n console.log(` ✅ /m/ 端点工作正常`)\n } else if (response.status === 307 || response.status === 302) {\n console.log(` ✅ /m/ 端点返回重定向（正常）`)\n console.log(` Location: ${response.headers.get('location')?.substring(0, 80)}...`)\n } else if (response.status === 404) {\n console.log(` ❌ MediaObject 未找到（404）`)\n } else {\n console.log(` ⚠️ 状态码: ${response.status}`)\n }\n } catch (error) {\n console.log(` ⚠️ 请求失败（可能服务器未启动）:`, error)\n }\n } else {\n console.log(` 跳过测试（URL 格式不正确）`)\n }\n } catch (error) {\n console.log(` 跳过测试:`, error)\n }\n\n // 清理\n console.log('\\n7️⃣ 清理测试数据:')\n try {\n const { deleteObject } = await import('../src/lib/storage')\n await deleteObject(storedKey)\n console.log(` ✅ 删除成功`)\n } catch (error) {\n console.log(` ⚠️ 删除失败:`, error)\n }\n\n console.log('\\n✨ 测试完成!')\n}\n\ntestFullImageFlow().catch(console.error)\n"
},
{
"path": "scripts/test-image-url-flow.ts",
"content": "/**\n * 图片 URL 流程测试\n * 模拟图片生成后的存储和读取流程\n * 运行: npx tsx scripts/test-image-url-flow.ts\n */\nimport { config } from 'dotenv'\nconfig()\n\nimport { uploadObject, getSignedUrl, extractStorageKey, toFetchableUrl } from '../src/lib/storage'\nimport { keyToSignedUrl, addSignedUrlToLocation } from '../src/lib/storage'\nimport { encodeImageUrls, decodeImageUrlsFromDb } from '../src/lib/contracts/image-urls-contract'\nimport { randomUUID } from 'crypto'\n\nasync function testImageUrlFlow() {\n console.log('🧪 测试图片 URL 全流程...\\n')\n\n // 1. 模拟上传图片到存储\n console.log('1️⃣ 模拟上传图片:')\n const testKey = `images/location-${randomUUID()}.jpg`\n const testImageContent = Buffer.from('fake-image-data')\n\n let storedKey: string\n try {\n storedKey = await uploadObject(testImageContent, testKey)\n console.log(` ✅ 上传成功，返回 key: ${storedKey}`)\n } catch (error) {\n console.log(` ❌ 上传失败:`, error)\n process.exit(1)\n }\n\n // 2. 模拟存储到数据库（encodeImageUrls）\n console.log('\\n2️⃣ 模拟数据库存储（encodeImageUrls）:')\n const imageUrlsArray = [storedKey]\n const dbValue = encodeImageUrls(imageUrlsArray)\n console.log(` ✅ 数据库值: ${dbValue}`)\n\n // 3. 模拟从数据库读取（decodeImageUrlsFromDb）\n console.log('\\n3️⃣ 模拟数据库读取（decodeImageUrlsFromDb）:')\n const decodedKeys = decodeImageUrlsFromDb(dbValue)\n console.log(` ✅ 解析出的 keys: ${JSON.stringify(decodedKeys)}`)\n\n // 4. 测试 keyToSignedUrl（用于 API 返回给前端）\n console.log('\\n4️⃣ 测试 keyToSignedUrl（API 层转换）:')\n for (const key of decodedKeys) {\n const signedUrl = keyToSignedUrl(key)\n console.log(` Key: ${key}`)\n console.log(` → Signed URL: ${signedUrl}`)\n\n // 检查是否是 /api/storage/sign 格式\n if (signedUrl?.startsWith('/api/storage/sign')) {\n console.log(` ✅ 正确生成了签名 URL 路径`)\n } else if (signedUrl?.startsWith('http')) {\n console.log(` ⚠️ 返回了完整 HTTP URL，可能无法直接访问`)\n } else {\n console.log(` ⚠️ URL 格式: ${signedUrl}`)\n }\n }\n\n // 5. 测试 addSignedUrlToLocation（完整对象转换）\n console.log('\\n5️⃣ 测试 addSignedUrlToLocation（完整对象转换）:')\n const mockLocationFromDb = {\n id: 'loc-123',\n name: '测试场景',\n images: [\n {\n id: 'img-1',\n imageUrl: storedKey,\n imageIndex: 0,\n }\n ]\n }\n\n const locationWithSignedUrls = addSignedUrlToLocation(mockLocationFromDb)\n console.log(` 转换后的 location.images:`)\n for (const img of locationWithSignedUrls.images || []) {\n console.log(` - imageIndex: ${img.imageIndex}`)\n console.log(` - imageUrl: ${img.imageUrl}`)\n\n if (img.imageUrl?.startsWith('/api/storage/sign')) {\n console.log(` ✅ 正确: 是相对路径签名 URL`)\n } else if (img.imageUrl?.startsWith('http://127.0.0.1:19000')) {\n console.log(` ❌ 错误: 是 MinIO 直链，可能需要签名`)\n } else if (img.imageUrl?.startsWith('http')) {\n console.log(` ⚠️ 是外部 HTTP URL`)\n } else {\n console.log(` ⚠️ 其他格式: ${img.imageUrl}`)\n }\n }\n\n // 6. 测试 getSignedUrl 直接调用\n console.log('\\n6️⃣ 测试 getSignedUrl 直接调用:')\n const directSignedUrl = getSignedUrl(storedKey)\n console.log(` Key: ${storedKey}`)\n console.log(` → URL: ${directSignedUrl}`)\n\n // 7. 测试 extractStorageKey\n console.log('\\n7️⃣ 测试 extractStorageKey（从各种 URL 提取 key）:')\n const testUrls = [\n storedKey,\n `http://127.0.0.1:19000/waoowaoo/${storedKey}`,\n directSignedUrl,\n ]\n for (const url of testUrls) {\n const extracted = extractStorageKey(url)\n console.log(` ${url.substring(0, 60)}...`)\n console.log(` → extracted: ${extracted}`)\n }\n\n // 8. 清理测试数据\n console.log('\\n8️⃣ 清理测试数据:')\n try {\n const { deleteObject } = await import('../src/lib/storage')\n await deleteObject(storedKey)\n console.log(` ✅ 删除成功`)\n } catch (error) {\n console.log(` ⚠️ 删除失败（可忽略）:`, error)\n }\n\n console.log('\\n✨ 测试完成!')\n console.log('\\n📋 总结:')\n console.log(' 如果第4、5步返回的是 /api/storage/sign?key=... 格式 → ✅ 正常')\n console.log(' 如果第4、5步返回的是 http://127.0.0.1:19000/... 格式 → ❌ 需要修复')\n}\n\ntestImageUrlFlow().catch(console.error)\n"
},
{
"path": "scripts/test-minio.ts",
"content": "/**\n * MinIO 存储测试脚本\n * 运行: npx tsx scripts/test-minio.ts\n */\nimport { config } from 'dotenv'\nconfig() // 加载 .env 文件\n\nimport { getStorageProvider, uploadObject, getSignedObjectUrl, getObjectBuffer, deleteObject } from '../src/lib/storage'\nimport { randomUUID } from 'crypto'\n\nasync function testMinio() {\n console.log('🧪 开始测试 MinIO 存储...\\n')\n\n // 1. 检查环境变量\n console.log('1️⃣ 检查环境变量:')\n const requiredEnv = [\n 'STORAGE_TYPE',\n 'MINIO_ENDPOINT',\n 'MINIO_ACCESS_KEY',\n 'MINIO_SECRET_KEY',\n 'MINIO_BUCKET',\n ]\n for (const key of requiredEnv) {\n const value = process.env[key]\n if (value) {\n // 隐藏敏感信息\n const displayValue = key.includes('SECRET') || key.includes('KEY') && key !== 'STORAGE_TYPE'\n ? '*'.repeat(Math.min(value.length, 8))\n : value\n console.log(` ✅ ${key}=${displayValue}`)\n } else {\n console.log(` ❌ ${key}=未设置`)\n }\n }\n\n // 2. 初始化 Provider\n console.log('\\n2️⃣ 初始化存储 Provider:')\n try {\n const provider = getStorageProvider()\n console.log(` ✅ Provider 类型: ${provider.kind}`)\n } catch (error) {\n console.log(` ❌ 初始化失败:`, error)\n process.exit(1)\n }\n\n // 3. 测试上传\n console.log('\\n3️⃣ 测试上传:')\n const testKey = `test/${randomUUID()}.txt`\n const testContent = `Hello MinIO! 测试时间: ${new Date().toISOString()}`\n let uploadedKey: string\n\n try {\n uploadedKey = await uploadObject(Buffer.from(testContent), testKey)\n console.log(` ✅ 上传成功: ${uploadedKey}`)\n } catch (error) {\n console.log(` ❌ 上传失败:`, error)\n process.exit(1)\n }\n\n // 4. 测试获取签名 URL\n console.log('\\n4️⃣ 测试获取签名 URL:')\n let signedUrl: string\n try {\n signedUrl = await getSignedObjectUrl(uploadedKey, 300)\n console.log(` ✅ 签名 URL 生成成功`)\n console.log(` URL: ${signedUrl.substring(0, 100)}...`)\n } catch (error) {\n console.log(` ❌ 签名 URL 生成失败:`, error)\n process.exit(1)\n }\n\n // 5. 测试下载\n console.log('\\n5️⃣ 测试下载:')\n try {\n const buffer = await getObjectBuffer(uploadedKey)\n const content = buffer.toString()\n if (content === testContent) {\n console.log(` ✅ 下载成功，内容匹配`)\n } else {\n console.log(` ❌ 下载成功，但内容不匹配`)\n console.log(` 预期: ${testContent}`)\n console.log(` 实际: ${content}`)\n }\n } catch (error) {\n console.log(` ❌ 下载失败:`, error)\n process.exit(1)\n }\n\n // 6. 通过 HTTP 访问签名 URL\n console.log('\\n6️⃣ 测试通过 HTTP 访问签名 URL:')\n try {\n const response = await fetch(signedUrl)\n if (response.ok) {\n const content = await response.text()\n if (content === testContent) {\n console.log(` ✅ HTTP 访问成功，内容匹配`)\n } else {\n console.log(` ❌ HTTP 访问成功，但内容不匹配`)\n }\n } else {\n console.log(` ❌ HTTP 访问失败: ${response.status} ${response.statusText}`)\n }\n } catch (error) {\n console.log(` ❌ HTTP 请求失败:`, error)\n }\n\n // 7. 清理测试文件\n console.log('\\n7️⃣ 清理测试文件:')\n try {\n await deleteObject(uploadedKey)\n console.log(` ✅ 删除成功`)\n } catch (error) {\n console.log(` ❌ 删除失败:`, error)\n }\n\n console.log('\\n✨ 测试完成!')\n}\n\ntestMinio().catch(console.error)\n"
},
{
"path": "scripts/test-regression-runner.sh",
"content": "#!/usr/bin/env bash\nset -euo pipefail\n\nif [ \"$#\" -eq 0 ]; then\n echo \"[regression-runner] missing command\"\n exit 2\nfi\n\nLOG_FILE=\"$(mktemp -t regression-runner.XXXXXX.log)\"\n\nset +e\n\"$@\" 2>&1 | tee \"$LOG_FILE\"\nCMD_STATUS=${PIPESTATUS[0]}\nset -e\n\nif [ \"$CMD_STATUS\" -ne 0 ]; then\n echo\n echo \"[regression-runner] regression failed, collecting diagnostics...\"\n\n FAILED_FILES=\"$(grep -E '^ FAIL ' \"$LOG_FILE\" | sed -E 's/^ FAIL ([^ ]+).*/\\1/' | sort -u || true)\"\n if [ -z \"$FAILED_FILES\" ]; then\n echo \"[regression-runner] no explicit FAIL file lines found in output\"\n else\n echo \"[regression-runner] failed files:\"\n while IFS= read -r file; do\n [ -z \"$file\" ] && continue\n echo \" - $file\"\n LAST_COMMIT=\"$(git log -n 1 --format='%h %ad %an %s' --date=short -- \"$file\" || true)\"\n FIRST_COMMIT=\"$(git log --diff-filter=A --follow --format='%h %ad %an %s' --date=short -- \"$file\" | tail -n 1 || true)\"\n if [ -n \"$LAST_COMMIT\" ]; then\n echo \" latest: $LAST_COMMIT\"\n fi\n if [ -n \"$FIRST_COMMIT\" ]; then\n echo \" first: $FIRST_COMMIT\"\n fi\n done <<< \"$FAILED_FILES\"\n fi\nfi\n\nrm -f \"$LOG_FILE\"\nexit \"$CMD_STATUS\"\n"
},
{
"path": "scripts/test-sign-api.ts",
"content": "/**\n * 测试 /api/storage/sign 端点\n * 运行: npx tsx scripts/test-sign-api.ts\n */\nimport { config } from 'dotenv'\nconfig()\n\nimport { uploadObject, getSignedObjectUrl } from '../src/lib/storage'\nimport { randomUUID } from 'crypto'\nimport http from 'http'\n\nasync function testSignApi() {\n console.log('🧪 测试 /api/storage/sign API...\\n')\n\n // 1. 上传测试文件\n console.log('1️⃣ 上传测试文件:')\n const testKey = `images/test-${randomUUID()}.txt`\n const testContent = 'Hello from MinIO test!'\n \n await uploadObject(Buffer.from(testContent), testKey)\n console.log(` ✅ 上传成功: ${testKey}`)\n\n // 2. 生成签名 URL（服务端直接调用）\n console.log('\\n2️⃣ 服务端生成签名 URL:')\n const signedUrl = await getSignedObjectUrl(testKey, 300)\n console.log(` URL: ${signedUrl}`)\n\n // 3. 测试直接访问签名 URL\n console.log('\\n3️⃣ 测试直接访问签名 URL:')\n try {\n const response = await fetch(signedUrl)\n if (response.ok) {\n const content = await response.text()\n console.log(` ✅ 访问成功，内容: \"${content}\"`)\n } else {\n console.log(` ❌ 访问失败: ${response.status} ${response.statusText}`)\n }\n } catch (error) {\n console.log(` ❌ 请求失败:`, error)\n }\n\n // 4. 测试 /api/storage/sign 端点（模拟前端访问）\n console.log('\\n4️⃣ 测试 /api/storage/sign 端点（模拟前端）:')\n const signApiUrl = `http://localhost:3000/api/storage/sign?key=${encodeURIComponent(testKey)}&expires=300`\n console.log(` URL: ${signApiUrl}`)\n \n try {\n const response = await fetch(signApiUrl, { redirect: 'manual' })\n console.log(` 状态: ${response.status}`)\n console.log(` Location: ${response.headers.get('location')}`)\n \n if (response.status === 307 || response.status === 302) {\n const redirectUrl = response.headers.get('location')\n console.log(` ✅ 重定向 URL: ${redirectUrl?.substring(0, 80)}...`)\n \n // 5. 测试跟随重定向\n console.log('\\n5️⃣ 测试跟随重定向访问图片:')\n const finalResponse = await fetch(signApiUrl, { redirect: 'follow' })\n if (finalResponse.ok) {\n const content = await finalResponse.text()\n console.log(` ✅ 最终访问成功，内容: \"${content}\"`)\n } else {\n console.log(` ❌ 最终访问失败: ${finalResponse.status}`)\n }\n } else {\n const body = await response.text()\n console.log(` 响应: ${body.substring(0, 200)}`)\n }\n } catch (error) {\n console.log(` ❌ 请求失败（可能服务器未启动）:`, error)\n }\n\n // 6. 测试 /api/cos/image 端点（旧版兼容）\n console.log('\\n6️⃣ 测试 /api/cos/image 端点（旧版兼容）:')\n const cosApiUrl = `http://localhost:3000/api/cos/image?key=${encodeURIComponent(testKey)}&expires=300`\n console.log(` URL: ${cosApiUrl}`)\n \n try {\n const response = await fetch(cosApiUrl, { redirect: 'manual' })\n console.log(` 状态: ${response.status}`)\n console.log(` Location: ${response.headers.get('location')}`)\n } catch (error) {\n console.log(` ❌ 请求失败（可能服务器未启动）:`, error)\n }\n\n // 清理\n console.log('\\n7️⃣ 清理测试文件:')\n const { deleteObject } = await import('../src/lib/storage')\n await deleteObject(testKey)\n console.log(` ✅ 清理完成`)\n\n console.log('\\n✨ 测试完成!')\n}\n\ntestSignApi().catch(console.error)\n"
},
{
"path": "scripts/tmp-cleanup-project-models.mjs",
"content": "import { PrismaClient } from '@prisma/client';\nconst p = new PrismaClient();\nsetTimeout(() => { console.error('TIMEOUT'); process.exit(1); }, 30000);\n\nconst userId = '3d84c341-87d7-4165-971d-a3f6c576aa21';\nconst needle = 'gemini-compatible:5b127c32-136e-4e5a-af74-8bae3e28be7a';\nconst modelFields = ['characterModel', 'locationModel', 'storyboardModel', 'editModel'];\n\n// novelPromotionData is a relation, query directly\nconst npProjects = await p.novelPromotionProject.findMany({\n where: { project: { userId } },\n select: { id: true, projectId: true, characterModel: true, locationModel: true, storyboardModel: true, editModel: true, project: { select: { name: true } } }\n});\n\nlet totalCleaned = 0;\n\nfor (const np of npProjects) {\n const updates = {};\n const cleanedFields = [];\n\n for (const field of modelFields) {\n if (typeof np[field] === 'string' && np[field].includes(needle)) {\n updates[field] = '';\n cleanedFields.push(`${field}: ${np[field]}`);\n }\n }\n\n if (cleanedFields.length > 0) {\n await p.novelPromotionProject.update({\n where: { id: np.id },\n data: updates\n });\n console.log(`✓ ${np.project.name} (${np.projectId}): cleared ${cleanedFields.length} fields`);\n cleanedFields.forEach(f => console.log(` - ${f}`));\n totalCleaned++;\n }\n}\n\nconsole.log(`\\nDone. Cleaned ${totalCleaned} projects.`);\nawait p.$disconnect();\nprocess.exit(0);\n"
},
{
"path": "scripts/tmp-find-old-model.mjs",
"content": "import { PrismaClient } from '@prisma/client';\nconst p = new PrismaClient();\nsetTimeout(() => { console.error('TIMEOUT'); process.exit(1); }, 15000);\n\nconst userId = '3d84c341-87d7-4165-971d-a3f6c576aa21';\nconst needle = 'gemini-compatible:5b';\n\n// 1. Check userPreference default models\nconst pref = await p.userPreference.findUnique({\n where: { userId },\n select: { analysisModel: true, characterModel: true, locationModel: true, storyboardModel: true, editModel: true, videoModel: true }\n});\nconsole.log('=== UserPreference defaults ===');\nlet found = false;\nfor (const [k, v] of Object.entries(pref || {})) {\n if (typeof v === 'string' && v.includes(needle)) {\n console.log(' FOUND in', k, ':', v);\n found = true;\n }\n}\nif (!found) console.log(' (clean)');\n\n// 2. Check novelPromotionData JSON for any reference\nconst projects = await p.project.findMany({\n where: { userId },\n select: { id: true, name: true, novelPromotionData: true }\n});\nconsole.log('\\n=== Project novelPromotionData ===');\nfor (const proj of projects) {\n const data = JSON.stringify(proj.novelPromotionData || {});\n if (data.includes(needle)) {\n // Find which keys reference it\n const parsed = proj.novelPromotionData;\n for (const [k, v] of Object.entries(parsed || {})) {\n if (typeof v === 'string' && v.includes(needle)) {\n console.log(' FOUND in project', proj.id, '(' + proj.name + ') field:', k, '=', v);\n }\n }\n }\n}\n\nawait p.$disconnect();\nprocess.exit(0);\n"
},
{
"path": "scripts/watchdog.ts",
"content": "import { createScopedLogger } from '@/lib/logging/core'\nimport { prisma } from '@/lib/prisma'\nimport { addTaskJob } from '@/lib/task/queues'\nimport { resolveTaskLocaleFromBody } from '@/lib/task/resolve-locale'\nimport { markTaskFailed } from '@/lib/task/service'\nimport { publishTaskEvent } from '@/lib/task/publisher'\nimport { TASK_EVENT_TYPE, TASK_TYPE, type TaskType } from '@/lib/task/types'\nimport { cleanupAllProjectLogs } from '@/lib/logging/file-writer'\n\nconst INTERVAL_MS = Number.parseInt(process.env.WATCHDOG_INTERVAL_MS || '30000', 10) || 30000\nconst HEARTBEAT_TIMEOUT_MS = Number.parseInt(process.env.TASK_HEARTBEAT_TIMEOUT_MS || '90000', 10) || 90000\nconst TASK_TYPE_SET: ReadonlySet = new Set(Object.values(TASK_TYPE))\n// 每小时执行一次日志清理\nconst LOG_CLEANUP_INTERVAL_TICKS = Math.ceil(3600_000 / INTERVAL_MS)\nlet tickCount = 0\nconst logger = createScopedLogger({\n module: 'watchdog',\n action: 'watchdog.tick',\n})\n\nfunction toTaskType(value: string): TaskType | null {\n if (TASK_TYPE_SET.has(value)) {\n return value as TaskType\n }\n return null\n}\n\nfunction toTaskPayload(value: unknown): Record | null {\n if (value && typeof value === 'object' && !Array.isArray(value)) {\n return value as Record\n }\n return null\n}\n\nasync function recoverQueuedTasks() {\n const rows = await prisma.task.findMany({\n where: {\n status: 'queued',\n enqueuedAt: null,\n },\n take: 100,\n orderBy: { createdAt: 'asc' },\n })\n\n for (const task of rows) {\n const taskType = toTaskType(task.type)\n if (!taskType) {\n logger.error({\n action: 'watchdog.reenqueue_invalid_type',\n message: `invalid task type: ${task.type}`,\n taskId: task.id,\n projectId: task.projectId,\n userId: task.userId,\n errorCode: 'INVALID_PARAMS',\n retryable: false,\n })\n continue\n }\n try {\n const locale = resolveTaskLocaleFromBody(task.payload)\n if (!locale) {\n await markTaskFailed(task.id, 'TASK_LOCALE_REQUIRED', 'task locale is missing')\n logger.error({\n action: 'watchdog.reenqueue_locale_missing',\n message: 'task locale is missing',\n taskId: task.id,\n projectId: task.projectId,\n userId: task.userId,\n errorCode: 'TASK_LOCALE_REQUIRED',\n retryable: false,\n })\n continue\n }\n\n await addTaskJob({\n taskId: task.id,\n type: taskType,\n locale,\n projectId: task.projectId,\n episodeId: task.episodeId,\n targetType: task.targetType,\n targetId: task.targetId,\n payload: toTaskPayload(task.payload),\n userId: task.userId,\n })\n await prisma.task.update({\n where: { id: task.id },\n data: {\n enqueuedAt: new Date(),\n enqueueAttempts: { increment: 1 },\n lastEnqueueError: null,\n },\n })\n logger.info({\n action: 'watchdog.reenqueue',\n message: 'watchdog re-enqueued queued task',\n taskId: task.id,\n projectId: task.projectId,\n userId: task.userId,\n details: {\n type: task.type,\n targetType: task.targetType,\n targetId: task.targetId,\n },\n })\n } catch (error: unknown) {\n const message = error instanceof Error ? error.message : 're-enqueue failed'\n await prisma.task.update({\n where: { id: task.id },\n data: {\n enqueueAttempts: { increment: 1 },\n lastEnqueueError: message,\n },\n })\n logger.error({\n action: 'watchdog.reenqueue_failed',\n message,\n taskId: task.id,\n projectId: task.projectId,\n userId: task.userId,\n errorCode: 'EXTERNAL_ERROR',\n retryable: true,\n })\n }\n }\n}\n\nasync function cleanupZombieProcessingTasks() {\n const timeoutAt = new Date(Date.now() - HEARTBEAT_TIMEOUT_MS)\n const rows = await prisma.task.findMany({\n where: {\n status: 'processing',\n heartbeatAt: { lt: timeoutAt },\n },\n take: 100,\n })\n\n for (const task of rows) {\n if ((task.attempt || 0) >= (task.maxAttempts || 5)) {\n await markTaskFailed(task.id, 'WATCHDOG_TIMEOUT', 'Task heartbeat timeout')\n await publishTaskEvent({\n taskId: task.id,\n projectId: task.projectId,\n userId: task.userId,\n type: TASK_EVENT_TYPE.FAILED,\n payload: { reason: 'watchdog_timeout' },\n })\n logger.error({\n action: 'watchdog.fail_timeout',\n message: 'watchdog marked task as failed due to heartbeat timeout',\n taskId: task.id,\n projectId: task.projectId,\n userId: task.userId,\n errorCode: 'WATCHDOG_TIMEOUT',\n retryable: true,\n })\n continue\n }\n\n await prisma.task.update({\n where: { id: task.id },\n data: {\n status: 'queued',\n enqueuedAt: null,\n heartbeatAt: null,\n startedAt: null,\n },\n })\n await publishTaskEvent({\n taskId: task.id,\n projectId: task.projectId,\n userId: task.userId,\n type: TASK_EVENT_TYPE.CREATED,\n payload: { reason: 'watchdog_requeue' },\n })\n logger.warn({\n action: 'watchdog.requeue_processing',\n message: 'watchdog re-queued stalled processing task',\n taskId: task.id,\n projectId: task.projectId,\n userId: task.userId,\n retryable: true,\n })\n }\n}\n\nasync function tick() {\n tickCount++\n const startedAt = Date.now()\n try {\n await recoverQueuedTasks()\n await cleanupZombieProcessingTasks()\n // 每小时清理一次日志（过滤 24h 前内容）\n if (tickCount % LOG_CLEANUP_INTERVAL_TICKS === 0) {\n void cleanupAllProjectLogs()\n }\n logger.info({\n action: 'watchdog.tick.ok',\n message: 'watchdog tick completed',\n durationMs: Date.now() - startedAt,\n })\n } catch (error: unknown) {\n const message = error instanceof Error ? error.message : 'watchdog tick failed'\n logger.error({\n action: 'watchdog.tick.failed',\n message,\n durationMs: Date.now() - startedAt,\n errorCode: 'INTERNAL_ERROR',\n retryable: true,\n })\n }\n}\n\nlogger.info({\n action: 'watchdog.started',\n message: 'watchdog started',\n details: {\n intervalMs: INTERVAL_MS,\n heartbeatTimeoutMs: HEARTBEAT_TIMEOUT_MS,\n },\n})\nvoid tick()\nsetInterval(() => {\n void tick()\n}, INTERVAL_MS)\n"
},
{
"path": "src/app/[locale]/auth/signin/page.tsx",
"content": "'use client'\n\nimport { useState } from \"react\"\nimport { signIn } from \"next-auth/react\"\nimport { useTranslations } from 'next-intl'\nimport Navbar from \"@/components/Navbar\"\nimport { Link, useRouter } from '@/i18n/navigation'\n\nexport default function SignIn() {\n const [username, setUsername] = useState(\"\")\n const [password, setPassword] = useState(\"\")\n const [loading, setLoading] = useState(false)\n const [error, setError] = useState(\"\")\n const router = useRouter()\n const t = useTranslations('auth')\n\n const handleSubmit = async (e: React.FormEvent) => {\n e.preventDefault()\n setLoading(true)\n setError(\"\")\n\n try {\n const result = await signIn(\"credentials\", {\n username,\n password,\n redirect: false,\n })\n\n if (result?.error === 'RateLimited') {\n setError(t('rateLimited'))\n } else if (result?.error) {\n setError(t('loginFailed'))\n } else {\n router.push({ pathname: '/' })\n router.refresh()\n }\n } catch {\n setError(t('loginError'))\n } finally {\n setLoading(false)\n }\n }\n\n return (\n \n
\n
\n
\n
\n
\n
\n {t('welcomeBack')}\n \n
{t('loginTo')}
\n
\n\n
\n\n
\n
\n {t('noAccount')}{\" \"}\n
\n
\n\n
\n
\n
\n
\n
\n
\n
\n
\n
\n
\n
\n
\n {t('createAccount')}\n \n
{t('joinPlatform')}
\n
\n\n
\n\n
\n
\n {t('hasAccount')}{\" \"}\n
\n
\n\n
\n
\n
\n
\n
\n
{\n const { locale } = await params;\n const t = await getTranslations({ locale, namespace: 'layout' })\n\n return {\n title: t('title'),\n description: t('description'),\n icons: {\n icon: '/logo.ico?v=2',\n shortcut: '/logo.ico?v=2',\n apple: '/logo.png?v=2',\n },\n };\n}\n\nexport function generateStaticParams() {\n return locales.map((locale) => ({ locale }));\n}\n\nexport default async function LocaleLayout({\n children,\n params,\n}: {\n children: React.ReactNode;\n params: Promise<{ locale: string }>;\n}) {\n const { locale } = await params;\n\n // 验证 locale 是否有效\n if (!locales.includes(locale as SupportedLocale)) {\n notFound();\n }\n\n // 获取翻译消息\n const messages = await getMessages();\n\n return (\n \n \n {process.env.NODE_ENV === \"development\" && (\n \n \n \n {children}\n \n \n\n \n ![\"🚀](\"images/cta-banner.png\")
\n \n![\"waoowaoo\"](\"public/banner.png\")
\n![](\"images/dab6b4105e3260f37ba2d5f536dce259.jpg\")
\n\n---\n\n## ✨ 功能特性\n\n- 🎬 **AI 剧本分析** — 自动解析小说，提取角色、场景、剧情\n- 🎨 **角色 & 场景生成** — AI 生成一致性人物和场景图片\n- 📽️ **分镜视频制作** — 自动生成分镜头并合成视频\n- 🎙️ **AI 配音** — 多角色语音合成\n- 🌐 **多语言支持** — 中文 / 英文界面，右上角一键切换\n\n---\n\n## 🚀 快速开始\n\n**前提条件**：安装 [Docker Desktop](https://docs.docker.com/get-docker/)\n\n### 方式一：拉取预构建镜像（最简单）\n\n无需克隆仓库，下载即用：\n\n```bash\n# 下载 docker-compose.yml\ncurl -O https://raw.githubusercontent.com/saturndec/waoowaoo/main/docker-compose.yml\n\n# 启动所有服务\ndocker compose up -d\n```\n\n> ⚠️ 当前为测试版，版本间数据库不兼容。升级请先清除旧数据：\n\n```bash\ndocker compose down -v\ndocker rmi ghcr.io/saturndec/waoowaoo:latest\ncurl -O https://raw.githubusercontent.com/saturndec/waoowaoo/main/docker-compose.yml\ndocker compose up -d\n```\n\n> 启动后请**清空浏览器缓存**并重新登录，避免旧版本缓存导致异常。\n\n### 方式二：克隆仓库 + Docker 构建（完全控制）\n\n```bash\ngit clone https://github.com/saturndec/waoowaoo.git\ncd waoowaoo\ndocker compose up -d\n```\n\n更新版本：\n```bash\ngit pull\ndocker compose down && docker compose up -d --build\n```\n\n### 方式三：本地开发模式（开发者）\n\n```bash\ngit clone https://github.com/saturndec/waoowaoo.git\ncd waoowaoo\n\n# 复制环境变量配置文件（必须在 npm install 之前完成）\ncp .env.example .env\n# ⚠️ 编辑 .env，填入你的 AI API Key（NEXTAUTH_URL 默认已是 http://localhost:3000，无需修改）\n\nnpm install\n\n# 只启动基础设施\n# 注意：docker-compose.yml 将服务映射到非标准端口，.env.example 已按此预设\nmysql:13306 redis:16379 minio:19000\ndocker compose up mysql redis minio -d\n\n# 初始化数据库表结构（首次必须执行，跳过会导致启动后报错）\nnpx prisma db push\n\n# 启动开发服务器\nnpm run dev\n```\n\n> [!WARNING]\n> 跳过 `npx prisma db push` 会导致所有数据库表不存在，启动后报错 `The table 'tasks' does not exist`。请务必先运行此命令再启动开发服务器。\n\n---\n\n访问 [http://localhost:13000](http://localhost:13000)（方式一、二）或 [http://localhost:3000](http://localhost:3000)（方式三）开始使用！\n\n> 首次启动会自动完成数据库初始化，无需任何额外配置。\n\n> [!TIP]\n> **如果遇到网页卡顿**：HTTP 模式下浏览器可能限制并发连接。可安装 [Caddy](https://caddyserver.com/docs/install) 启用 HTTPS：\n> ```bash\n> caddy run --config Caddyfile\n> ```\n> 然后访问 [https://localhost:1443](https://localhost:1443)\n\n---\n\n## 🔧 API 配置\n\n启动后进入**设置中心**配置 AI 服务的 API Key，内置配置教程。\n\n> 💡 **注意**：目前仅推荐使用各服务商官方 API，第三方兼容格式（OpenAI Compatible）尚不完善，后续版本会持续优化。\n\n---\n\n## 📦 技术栈\n\n- **框架**: Next.js 15 + React 19\n- **数据库**: MySQL + Prisma ORM\n- **队列**: Redis + BullMQ\n- **样式**: Tailwind CSS v4\n- **认证**: NextAuth.js\n\n---\n\n## 📦 页面功能预览\n\n\n\n\n\n\n---\n\n## 🤝 参与方式\n\n本项目由核心团队独立维护。欢迎你通过以下方式参与：\n\n- 🐛 提交 [Issue](https://github.com/saturndec/waoowaoo/issues) 反馈 Bug\n- 💡 提交 [Issue](https://github.com/saturndec/waoowaoo/issues) 提出功能建议\n- 🔧 提交 Pull Request 供参考 — 我们会认真审阅每一个 PR 的思路，但最终由团队自行实现修复，不会直接合并外部 PR\n\n---\n\n**Made with ❤️ by waoowaoo team**\n\n## Star History\n\n[](https://www.star-history.com/#saturndec/waoowaoo&type=date&legend=top-left)\n"
},
{
"path": "README_en.md",
"content": "