Doc Search

About

Empower your AI models with accurate, up-to-date information by leveraging this high-performance Model Context Protocol (MCP) server. Built with Python, it efficiently fetches and transforms official library documentation from various sources into pristine, Large Language Model-friendly Markdown. By stripping irrelevant content like headers, footers, and sidebars, and formatting it intelligently, it ensures LLMs receive only the essential technical context for better understanding and generation. This tool is invaluable for enhancing AI-driven knowledge retrieval, supporting both pre-configured and dynamically registered documentation sources.

Key Features

• Parallel Scraping using asyncio and httpx for fast fetching.
• Dynamic Registry supporting pre-configured and runtime registration of documentation sources.
• Auto-Discovery of documentation URLs via DuckDuckGo for unregistered libraries.
• Smart Parsing that strips headers, footers, and sidebars for relevant content.
• 1 GitHub stars
• Markdown Conversion from HTML into structured, LLM-friendly Markdown.

Use Cases

• Integrating with LLM applications like Claude Desktop to provide real-time, context-aware documentation.
• Enhancing AI model responses with accurate and current technical information from library documentation.
• Dynamically adding new software library documentation sources for immediate search and retrieval by MCP clients.