LLMDoc Documentation Search

Example Use Cases

• 1

  Semantically search multiple LLM-friendly documentation sources.

• 2

  Up-to-date, indexed documentation with source filtering and attribution.

• 3

  Retrieve full or excerpted document content with pagination support.

Description

LLMDoc indexes and semantically searches documentation from llms.txt files using a hybrid two-stage search combining DuckDB full-text search and BM25 reranking. It supports named sources, persistent DuckDB storage, and automatic background refresh to keep documentation up-to-date. The server integrates with MCP clients to enable fast, precise retrieval of relevant documentation passages with source attribution.