Documentation Retriever

Acerca de

This tool provides an asynchronous server that exposes a powerful `get_docs` function. It performs targeted web searches within specified official documentation domains (e.g., uv, LangChain, OpenAI), fetches the relevant web pages, and then intelligently cleans the raw HTML into concise, readable text chunks using a large language model. Designed to enhance Retrieval-Augmented Generation (RAG) systems, it meticulously labels each content block with its original source URL, ensuring full traceability for the information provided to an LLM.

Características Principales

• Search restricted to official documentation domains (e.g., uv, LangChain, OpenAI, Llama-Index).
• 1 GitHub stars
• Tool: `get_docs(query, library)` returns concatenated cleaned sections with `SOURCE:` labels.
• Streaming-safe async design, chunking large HTML pages before LLM cleaning.
• Utilizes Serper.dev for Google-like search and Groq API for fast LLM-based content cleaning.
• Includes a client example demonstrating integration for LLM-powered answer generation.

Casos de Uso

• Enabling AI agents to access and synthesize information directly from technical library documentation.
• Powering Retrieval-Augmented Generation (RAG) systems with up-to-date, official documentation.
• Automating the collection and preparation of structured knowledge from web-based documentation for analysis.