QMD Markdown Knowledge Search
Searches personal markdown knowledge bases and documentation using a local hybrid search engine combining BM25 and vector semantic search.
QMD is a specialized skill for Claude Code that enables lightning-fast, local-first search across your personal markdown notes, meeting transcripts, and project documentation. By leveraging a hybrid search approach—combining traditional keyword matching with semantic vector search and LLM-powered re-ranking—it allows users to find information using natural language queries even when exact keywords aren't present. It is an essential tool for developers and knowledge workers who need to instantly retrieve context from extensive local documentation or personal knowledge bases while maintaining data privacy.
Key Features
01Multi-collection management to organize notes, transcripts, and docs separately
02Local-first execution for data privacy and high-speed indexing
035,804 GitHub stars
04Hybrid search combining BM25 keyword matching and semantic vector search
05Support for document retrieval via doc IDs, file paths, or glob patterns
06LLM-powered result re-ranking for maximum search accuracy
Use Cases
01Searching personal knowledge bases to synthesize information for research tasks
02Finding specific information across thousands of meeting notes or transcripts
03Retrieving internal project documentation when exact keywords are forgotten