MCP Server Documentation

🚀 MCP Document Search Server

The MCP Document Search Server is a powerful tool built on FastMCP, enabling AI systems to intelligently search across the documentation of multiple popular frameworks and libraries. It ensures that AI models can quickly access and retrieve relevant information through a unified interface.

✨ Features

📚 Multi-library Support

Search through documentation across multiple libraries:

• LangChain - JavaScript/TypeScript LangChain documentation
• LangGraph - LangGraph.js documentation
• Next.js - Next.js framework documentation
• Tailwind CSS - Utility-first CSS framework
• FastMCP - FastMCP documentation
• Framer Motion - React animation library

🔍 Intelligent Search

• Handle library name variations intelligently.
• Provide high-precision DuckDuckGo search results.
• Offer site-specific search functionality.

⚡ Performance Optimization

• Implement asynchronous processing.
• Handle web requests efficiently.
• Load resources in parallel.

🛡️ Error Handling

• Feature an automatic retry mechanism.
• Control timeouts.
• Validate status codes.

📦 Installation Requirements

• Operating System: Supports Unix, Linux, and macOS. Windows requires additional configuration.
• Python Version: 3.8 or higher.
• Package Manager: It is recommended to use pip or uv package manager.
• Dependencies: The requests library needs to be installed.

🚀 Quick Start

1. Clone the repository:

git clone https://github.com/your-repository.git
cd mcp-docs-search-server

2. Set up a virtual environment (recommended):

python -m venv env
source env/bin/activate  # On Unix/Linux
# Or
.\env\Scripts\activate  # On Windows

3. Install dependencies:

pip install -r requirements.txt

4. Run the server:

python server.py

💻 Usage Examples

from mcp_search import search_document

# Search Framer Motion documentation
results = search_document("framer motion", "animations")
print(results)

# Search Tailwind CSS documentation
results = search_document("tailwind css", "utilities classes")
print(results)

📋 Testing Guide

Unit Testing

pytest tests/

Integration Testing

PYTEST_ARGS="-v" pytest integration_tests/

Asynchronous Testing

uvicorn async_server:app --reload

🗄 Project Structure

• mcp_search/: Core search module.
• server.py: Main server file.
• requirements.txt: Dependency management file.
• tests/: Unit test directory.

🔧 Configuration Guide

Supported Libraries

Define documentation URLs in config.py:

DOCS_URLS = {
    "new-library": "https://docs.new-library.com",
    # Existing entries...
}

Add aliases:

LIBRARY_ALIASES = {
    "new-lib": "new-library",
    # Existing entries...
}

Network Settings

Adjust timeouts and the number of results in config.py:

HTTP_TIMEOUT = 30.0        # Timeout in seconds
MAX_SEARCH_RESULTS = 2     # Number of results to retrieve

🤝 Contribution Guide

1. Fork the repository.
2. Create a feature branch.
3. Commit your changes.
4. Add tests for new features.
5. Ensure all tests pass.
6. Submit a pull request.

Steps to Add a New Library

1. Update DOCS_URLS and LIBRARY_ALIASES.
2. Test the integration.
3. Update the documentation.

🚨 Troubleshooting

• Timeout Errors: Increase the HTTP_TIMEOUT value.
• No Results: Try different search terms or verify the library name.
• Network Errors: Check your internet connection and the documentation URLs.

📄 License

This project is licensed under the MIT License. Please refer to the LICENSE file for details.

🙏 Acknowledgments

Thanks to the following projects and technologies:

• FastMCP for providing core functionality.
• DuckDuckGo for search support.
• pytest testing framework.