Context7 MCP Clone
Provides up-to-date, version-specific documentation and code examples for software libraries directly to LLMs, enabling resolution of library identifiers and retrieval of relevant documentation with code snippets.
README
Context7 MCP Clone - Up-to-Date Code Documentation for LLMs
A production-ready clone of Context7 MCP that provides up-to-date, version-specific documentation and code examples directly to LLMs and AI code editors.
🎯 Project Status
Phase 2 Complete: Core MCP Server and Backend API ready for development!
Completed Components
- ✅ pnpm monorepo with workspace configuration
- ✅ Docker Compose development and production environments
- ✅ PostgreSQL database schema with migrations
- ✅ NestJS backend API with authentication
- ✅ Library and documentation services
- ✅ MCP server with both tools (resolve-library-id, get-library-docs)
- ✅ Stdio and HTTP transports for MCP protocol
In Progress
- 🚀 Rate limiting with Redis
- 🚀 Crawler engine with GitHub API and web scraping
- 🚀 Markdown and HTML parsing with code extraction
- 🚀 Next.js web UI with Grafana theme
📦 Project Structure
atamai-mcp/
├── packages/
│ ├── mcp-server/ # MCP Protocol Server
│ ├── backend-api/ # NestJS REST API
│ ├── crawler-engine/ # Documentation Crawler
│ └── web-ui/ # Next.js Frontend
├── docker/ # Container configurations
├── scripts/ # Utility scripts
├── docker-compose.yml # Production compose
├── docker-compose.dev.yml # Development compose
├── PLAN.md # Detailed implementation plan
├── TODO.md # Task checklist
└── README.md # This file
🚀 Quick Start
Prerequisites
- Node.js 20+
- pnpm 8+
- Docker & Docker Compose
- PostgreSQL 16+ (or use docker-compose)
Development Setup
- Clone and install dependencies
cd /Users/atamai/develope/atamai-mcp
pnpm install
- Set up environment
cp .env.example .env
# Edit .env with your configuration
- Start development environment with Docker
docker-compose -f docker-compose.dev.yml up
This starts:
- PostgreSQL database on port 5432
- Redis cache on port 6379
- Backend API on port 5000
- MCP Server on port 3000
- Web UI on port 4000
- Run migrations
# Migrations run automatically when backend starts
# Or manually run:
cd packages/backend-api
pnpm run migration:run
📚 MCP Server
Tools Available
1. resolve-library-id
Converts a library name to Context7-compatible ID and returns ranked results.
Input:
{
"libraryName": "react"
}
Output:
{
"libraries": [
{
"id": "/facebook/react",
"name": "React",
"description": "A JavaScript library for building user interfaces",
"ecosystem": "javascript",
"stars": 220000,
"benchmarkScore": 98,
"reputation": "high",
"codeSnippets": 15420
}
],
"selected": "/facebook/react",
"reasoning": "Exact name match with highest benchmark score"
}
2. get-library-docs
Retrieves version-specific documentation and code examples.
Input:
{
"context7CompatibleLibraryID": "/facebook/react",
"topic": "hooks",
"page": 1,
"mode": "code"
}
Output:
{
"libraryId": "/facebook/react",
"library": {
"name": "React",
"full_name": "facebook/react",
"ecosystem": "javascript"
},
"version": "18.2.0",
"topic": "hooks",
"page": 1,
"totalPages": 5,
"documentation": [
{
"title": "Using the State Hook",
"type": "guide",
"content": "...",
"codeExamples": [
{
"language": "javascript",
"code": "const [count, setCount] = useState(0);",
"description": "Basic useState example"
}
],
"url": "https://react.dev/reference/react/useState"
}
]
}
Running MCP Server
Stdio Mode (for Claude Desktop, Cursor, etc.)
cd packages/mcp-server
pnpm start
# Or with hot reload:
pnpm dev
HTTP Mode
MCP_TRANSPORT=http MCP_SERVER_PORT=3000 pnpm start
# Access SSE endpoint at: http://localhost:3000/mcp/sse
🔌 Backend API
Authentication
Register
POST /api/v1/auth/register
Content-Type: application/json
{
"email": "user@example.com",
"password": "SecurePassword123!"
}
Login
POST /api/v1/auth/login
Content-Type: application/json
{
"email": "user@example.com",
"password": "SecurePassword123!"
}
Libraries
Search Libraries
GET /api/v1/libraries?query=react&ecosystem=javascript&page=1&limit=20
Authorization: Bearer <jwt_token>
Get Library by ID
GET /api/v1/libraries/{id}
Authorization: Bearer <jwt_token>
Get Available Ecosystems
GET /api/v1/libraries/ecosystems
Authorization: Bearer <jwt_token>
Documentation
Get Documentation
GET /api/v1/docs/{libraryId}?topic=hooks&page=1&mode=code
Authorization: Bearer <jwt_token>
Search Documentation
GET /api/v1/docs/search/{query}
Authorization: Bearer <jwt_token>
Swagger Documentation
Once the backend is running, visit:
http://localhost:5000/docs
📊 Database Schema
Core Tables
- users - User accounts with tier system
- api_keys - API key management with rate limits
- libraries - Library metadata
- library_versions - Version tracking
- documentation_pages - Documentation content with full-text search
- code_examples - Code snippets
- crawl_jobs - Crawler job tracking
- api_usage - Usage analytics
Features
- Full-text search on documentation
- Trigram similarity for fuzzy matching
- JSONB metadata support
- Automatic tsvector updates
- Materialized views for popular libraries
🔄 Deployment
Production Docker Compose
docker-compose up -d
Environment Variables
See .env.example for all available options:
DATABASE_URL- PostgreSQL connection stringREDIS_URL- Redis connection stringJWT_SECRET- JWT signing keyNODE_ENV- production/developmentAPI_PREFIX- API base path (default: /api/v1)
Kubernetes Deployment
Deploy manifests are in the plan for Phase 8.
📈 Development Roadmap
Phase 3: Crawler Engine (Weeks 5-6)
- [ ] BullMQ job queue setup
- [ ] GitHub crawler with Octokit
- [ ] Documentation site scraper
- [ ] Markdown/HTML parsers
- [ ] Code extraction engine
Phase 4: Web UI (Weeks 7-8)
- [ ] Next.js 15 setup with App Router
- [ ] Landing page with purple gradient theme
- [ ] Documentation browser
- [ ] Authentication pages
Phase 5: Dashboard & Admin (Weeks 9-10)
- [ ] User dashboard
- [ ] API key management
- [ ] Admin panel
- [ ] Analytics and charts
Phase 6-10: Completion & Launch
- [ ] Data seeding
- [ ] Testing & optimization
- [ ] Production deployment
- [ ] Beta & public launch
🧪 Testing
# Run tests for all packages
pnpm test
# Run tests with coverage
pnpm test:cov
# Watch mode
pnpm test:watch
📝 Code Style
# Format code
pnpm format
# Check formatting
pnpm format:check
# Lint
pnpm lint
# Type check
pnpm type-check
🐛 Debugging
Backend API
DEBUG=* pnpm dev
MCP Server
DEBUG=* pnpm dev
Docker Logs
docker-compose -f docker-compose.dev.yml logs -f [service-name]
📖 Documentation
- PLAN.md - Comprehensive implementation plan with architecture
- TODO.md - Detailed task checklist for all phases
- API Docs - Available at
http://localhost:5000/docs(Swagger)
🤝 Contributing
This is a solo development project. Please follow:
- TypeScript strict mode
- ESLint configuration
- Prettier formatting
- 80%+ test coverage
📄 License
MIT
🎯 Key Metrics (Target)
- Performance: MCP response <200ms (p95)
- Search: <100ms (p95)
- Uptime: 99.9%
- Coverage: 100+ libraries by month 6
- Users: 1000+ registered by month 3
Next Steps
- Run
docker-compose -f docker-compose.dev.yml up - Wait for migrations to complete
- Visit http://localhost:5000/docs for API documentation
- Start using the MCP server with Claude Desktop or Cursor
- Reference PLAN.md for detailed architecture and TODO.md for task tracking
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。