Provides a lightweight BM25 search server for efficiently querying content within Markdown files.
Sponsored
This tool offers a powerful and lightweight BM25 search server, designed to efficiently index and query content from a specified folder of Markdown files. It operates as an MCP server, exposing a `search` tool that allows seamless integration with clients supporting the Model Context Protocol. Beyond MCP integration, it also provides a versatile command-line interface for both interactive and single-query operations. Users can effortlessly add their `.md` documents to the `data/` directory, which are then automatically discovered and indexed for fast, accurate information retrieval using the configurable BM25Okapi algorithm.
Key Features
01Lightweight BM25 search server
02MCP server integration for contextual search
03CLI for interactive and single-query searching
04Automatic indexing of Markdown documents from a specified folder
05Configurable BM25 parameters (k1, b)
060 GitHub stars
Use Cases
01Developing and testing information retrieval systems with a simple yet effective BM25 implementation.
02Integrating contextual document search into MCP-compatible agents or applications.
03Quickly searching a local knowledge base or documentation maintained in Markdown.
