qmd

You can search meeting transcripts and markdown documentation using hybrid retrieval, which combines BM25 keyword search, vector search, and LLM reranking over local SQLite-backed indexes without cloud dependencies.

Yes, you can perform local AI semantic search without cloud dependencies because this Skill uses local SQLite-backed indexing to process your notes, project archives, and document collections securely.

Hybrid retrieval combines BM25 and vector search for personal knowledge bases by applying keyword search for exact terms, semantic vectors for conceptual matches, and LLM reranking to refine the final results.

Yes, MCP integration supports hands-free retrieval for automated workflows by allowing agents to query named collections and retrieve relevant information directly from local project archives and markdown documentation.

The best way to scope searches across multiple local document collections is to use collection-based retrieval with context descriptions, allowing you to query named collections of notes, docs, and transcripts accurately.

Yes, you can query project archives directly from the CLI because this Skill supports direct terminal use alongside MCP integration for searching local text collections and retrieving relevant passages.