QMD (Query Markup Documents)

QMD (Query Markup Documents) is a document format designed for AI-first knowledge management. It combines the readability of Markdown with structured metadata optimized for semantic search and AI retrieval systems.

QMD documents include: standard markdown content (human-readable text), frontmatter metadata (structured data about the document), query hints (suggested questions this content answers), semantic tags (topic and concept classifications), and relationship links (connections to related documents).

The format is designed to work well with RAG (Retrieval-Augmented Generation) systems by making it easier for embedding models and search engines to understand document structure, identify relevant sections, and retrieve the most appropriate content for user queries.

For AI knowledge bases, QMD offers advantages over plain text: better retrieval accuracy (metadata helps search engines find relevant content), document relationships (understanding how documents connect), query optimization (pre-defined questions improve matching), and structured sections (enabling more precise chunk retrieval).

While QMD is a niche format, the principles behind it — structured content, semantic metadata, and query optimization — are universally applicable to building effective AI knowledge bases on any platform.