docret-mcp-server
The Documentation Retrieval MCP Server is a tool for AI assistants to access up-to-date Python library documentation. It supports dynamic retrieval and parsing of documentation from several libraries, with plans to expand further. The project aims to ensure AI applications have access to the latest official documentation.
Documentation Retrieval MCP Server (DOCRET)
This project is a Model Context Protocol (MCP) server designed to help AI assistants access current documentation for various Python libraries, such as LangChain, LlamaIndex, and OpenAI. It enables dynamic documentation retrieval, asynchronous web searches through the SERPER API, and HTML parsing using BeautifulSoup. Users can easily extend the server to support additional libraries by updating configurations. The server provides an API for fetching documentation by querying relevant sources and processing the information.
Features
- Dynamic Documentation Retrieval
- Asynchronous Web Searches
- HTML Parsing
- Extensible Design
Prerequisites
- Python 3.8 or higher
- UV for Python Package Management
- Serper API key
- Claude Desktop or Code for testing
Installation
- Clone the repository
- Create and activate a virtual environment
- Install dependencies with pip or uv
Running the MCP Server
Setup required environment variables and start the server using python main.py.
Usage
API allows fetching documentation content by searching using specific topics and libraries.
Roadmap
- Add support for more libraries
- Implement caching
- Optimize web scraping
- Introduce REST API
- Add unit tests
- More MCP tools for development