LangChain Documentation MCP Server
A comprehensive dual-mode server that provides real-time access to LangChain documentation, API references, and code examples. Supports both FastAPI web service and native Model Context Protocol (MCP) server modes, fetching live data from official LangChain sources.
🚀 Features
- �️ Dual Server Modes - Run as FastAPI web service or native MCP server
- �📚 Live Documentation Search - Search through official LangChain documentation in real-time
- 🔍 API Reference Lookup - Get detailed API references from GitHub source code
- 🐙 GitHub Code Examples - Fetch real code examples from the LangChain repository
- 📖 Tutorial Discovery - Find and access LangChain tutorials and guides
- 📦 Version Tracking - Get latest version information from PyPI
- 🔗 Direct API Search - Search specifically through API reference documentation
- 🔌 MCP Protocol Support - Native Model Context Protocol implementation
🌐 Data Sources
This server fetches live data from:
- python.langchain.com - Official LangChain documentation
- GitHub LangChain Repository - Source code and examples
- PyPI - Latest version and release information
📋 API Endpoints
Core Endpoints
GET /- API documentation (Swagger UI)GET /health- Health check and service status
LangChain Documentation
GET /search- Search general documentationGET /search/api- Search API reference specificallyGET /api-reference/{class_name}- Get detailed API reference for a classGET /examples/github- Get real code examples from GitHubGET /tutorials- Get tutorials and guidesGET /latest-version- Get latest LangChain version info
🚀 Quick Start
Option 1: Docker Compose (Recommended)
-
Clone the repository
git clone https://github.com/LiteObject/langchain-mcp-server.git cd langchain-mcp-server -
Start the FastAPI server
docker-compose up --build -
Access the API
- API Documentation: http://localhost:8080/docs
- Health Check: http://localhost:8080/health
Option 2: Local Development
FastAPI Mode
-
Install dependencies
pip install -r requirements.txt -
Run the FastAPI server
# Using the main entry point python run.py # Or using the dedicated script python scripts/run_fastapi.py # Or directly with uvicorn uvicorn src.api.fastapi_app:app --host 0.0.0.0 --port 8000
MCP Server Mode
-
Install dependencies
pip install -r requirements.txt -
Run the MCP server
# Using the main entry point python run.py mcp # Or using the dedicated script python scripts/run_mcp.py
📚 Usage Examples
Search Documentation
# Search for "ChatOpenAI" in documentation
curl "http://localhost:8080/search?query=ChatOpenAI&limit=5"
# Search API reference specifically
curl "http://localhost:8080/search/api?query=embeddings"
Get API Reference
# Get detailed API reference for ChatOpenAI
curl "http://localhost:8080/api-reference/ChatOpenAI"
# Get API reference for LLMChain
curl "http://localhost:8080/api-reference/LLMChain"
Fetch Code Examples
# Get real examples from GitHub
curl "http://localhost:8080/examples/github?query=chatbot&limit=3"
# Get general examples
curl "http://localhost:8080/examples/github"
Get Tutorials
# Fetch all available tutorials
curl "http://localhost:8080/tutorials"
Version Information
# Get latest version from PyPI
curl "http://localhost:8080/latest-version"
🔌 MCP Server Usage
When running in MCP mode, the server provides the following tools:
Available MCP Tools
search_langchain_docs- Search LangChain documentationsearch_api_reference- Search API reference specificallyget_api_reference- Get detailed API reference for a classget_github_examples- Get code examples from GitHubget_tutorials- Get available tutorialsget_latest_version- Get latest LangChain version
MCP Client Integration
{
"mcpServers": {
"langchain-docs": {
"command": "python",
"args": ["path/to/langchain-mcp-server/run.py", "mcp"],
"env": {
"PYTHONPATH": "path/to/langchain-mcp-server"
}
}
}
}
🛠️ Configuration
Environment Variables
| Variable | Description | Default |
|---|---|---|
HOST | Server host address | 0.0.0.0 |
PORT | Server port | 8000 |
DEBUG | Enable debug mode | False |
LOG_LEVEL | Logging level | INFO |
REQUEST_TIMEOUT | Timeout for external API calls | 30 seconds |
GITHUB_TOKEN | GitHub API token (optional) | None |
Docker Configuration
The service runs on port 8080 by default to avoid conflicts. You can modify this in docker-compose.yml:
ports:
- "8080:8000" # Host:Container
🔧 Development
Project Structure
├── src/ # Main source code package
│ ├── main.py # Main entry point with dual mode support
│ ├── api/ # API layer
│ │ ├── fastapi_app.py # FastAPI application
│ │ └── mcp_server.py # Native MCP server implementation
│ ├── config/ # Configuration management
│ │ ├── settings.py # Application settings
│ │ └── logging.py # Logging configuration
│ ├── models/ # Data models and schemas
│ │ └── schemas.py # Pydantic models
│ ├── services/ # Business logic
│ │ └── langchain_service.py # LangChain documentation service
│ └── utils/ # Utility modules
│ ├── exceptions.py # Custom exceptions
│ └── helpers.py # Helper functions
├── scripts/ # Convenience scripts
│ ├── run_fastapi.py # Run FastAPI mode
│ ├── run_mcp.py # Run MCP mode
│ └── health_check.py # Health check utility
├── tests/ # Test suite
│ ├── test_api.py # API tests
│ ├── test_services.py # Service tests
│ └── test_integration.py # Integration tests
├── docs/ # Documentation
│ └── API.md # API documentation
├── logs/ # Log files
├── run.py # Simple entry point
├── requirements.txt # Python dependencies
├── pyproject.toml # Project configuration
├── Dockerfile # Docker configuration
├── docker-compose.yml # Docker Compose setup
├── DOCKER.md # Docker documentation
└── README.md # This file
Key Dependencies
- FastAPI - Web framework for REST API mode
- MCP - Native Model Context Protocol support
- FastAPI-MCP - MCP integration for FastAPI
- httpx - Async HTTP client for external API calls
- BeautifulSoup4 - HTML parsing for documentation scraping
- Pydantic - Data validation and settings management
- uvicorn - ASGI server for FastAPI
Adding New Endpoints
- Define Pydantic models for request/response
- Add endpoint function with proper type hints
- Include comprehensive docstrings
- Add error handling with specific exceptions
- Update health check endpoint count
🐛 Error Handling
The server includes robust error handling for:
- Network failures - Graceful degradation when external APIs are unavailable
- Rate limiting - Handles GitHub API rate limits
- Invalid requests - Proper HTTP status codes and error messages
- Timeouts - Configurable request timeouts
📊 Health Monitoring
The /health endpoint provides:
- Service status
- Available endpoints count
- Data source URLs
- Current timestamp
- Updated documentation sections
🔒 Security Considerations
- Rate Limiting - Consider implementing rate limiting for production
- CORS - Configure CORS headers if needed for web access
- API Keys - Add GitHub token for higher API limits
- Input Validation - All inputs are validated using Pydantic
🚀 Production Deployment
For production use, consider:
- Caching - Add Redis/Memcached for response caching
- Rate Limiting - Implement request rate limiting
- Monitoring - Add application monitoring and logging
- Load Balancing - Use multiple instances behind a load balancer
- Database - Store frequently accessed data
- CI/CD - Set up automated deployment pipeline
🤝 Contributing
- Fork the repository
- Create a feature branch
- Make your changes
- Add tests if applicable
- Submit a pull request
📄 License
This project is licensed under the MIT License - see the LICENSE file for details.
🔗 Related Links
- LangChain Documentation
- LangChain GitHub
- FastAPI Documentation
- Model Context Protocol
- MCP Python SDK
🆘 Support
If you encounter any issues:
- Check the health endpoint for service status (FastAPI mode)
- Review Docker logs:
docker-compose logs - Check application logs in the
logs/directory - Ensure network connectivity to external APIs
- Verify all dependencies are installed correctly
- For MCP mode issues, check the MCP client configuration
Note: This server requires internet connectivity to fetch live data from LangChain's official sources. API rate limits may apply for GitHub API calls.