Doctree Mcp

Doctree Mcp is an agentic document retrieval framework that enables AI agents to reason over and navigate structured document collections such as markdown, CSV, and JSONL files. Unlike embedding-based or similarity-search approaches, Doctree Mcp uses structural primitives—including a navigable tree, BM25 keyword search, glossary expansion, and row lookup—along with explicit agent skills to facilitate iterative, auditable content retrieval. It's particularly suitable for AI tools needing advanced, context-aware document search, wiki compilation, and progressive knowledge discovery over complex repositories.