jdocmunch-mcp

Documentation retrieval for AI agents: fetch the section you need, not the whole doc. Markdown, RST, AsciiDoc, Jupyter, HTML, OpenAPI and more. Runs locally, wo

What is jdocmunch-mcp?

Documentation retrieval for AI agents: fetch the section you need, not the whole doc. Markdown, RST, AsciiDoc, Jupyter, HTML, OpenAPI and more. Runs locally, works with Claude Code and any MCP client. Free for non-commercial use; commercial use needs a one-time license.

Overview of jdocmunch-mcp

What it is

jDocMunch is an MCP server for coding agents that retrieves the exact documentation section a task needs without loading whole files into the context window. It indexes a documentation set once by heading hierarchy, then fetches a single section, a heading subtree, or a ranked search result extracted byte-precisely from the original file. It runs locally and works with Claude Code and any MCP client.

Key features

  • Section-first retrieval: search and retrieve documentation by section, not just file path or keyword match.
  • Byte-precise extraction: full content is pulled on demand from exact byte offsets into the original file.
  • Stable section IDs: sections retain durable identities across re-indexing when path, heading text, and heading level remain unchanged.
  • Quality gate: every release runs a replay fixture over a frozen golden set and fails below nDCG 0.95.
  • Supports multiple documentation formats: Markdown, RST, AsciiDoc, Jupyter, HTML, OpenAPI and more.
  • Optional OCR support for scanned PDFs (requires an OCR engine already on the machine).

Use cases

  • AI coding agents that need to retrieve a specific documentation section (e.g., authentication configuration) without loading the entire file.
  • Reducing context window waste by fetching only the relevant heading subtree or ranked search result.
  • Works with any MCP-compatible client, including Claude Code.

Getting started

Requirements: Python 3.10+ and any MCP-compatible client. Install with uv: bash uv tool install jdocmunch-mcp jdocmunch-mcp init Alternative install paths: bash uvx jdocmunch-mcp pipx install jdocmunch-mcp pip install jdocmunch-mcp Verify: bash jdocmunch-mcp --version Manual Claude Code setup: bash claude mcp add -s user jdocmunch -- uvx jdocmunch-mcp Quickstart – index a local documentation folder: bash jdocmunch-mcp index-local --path ./docs Then inside the agent, use search_sections and get_section to retrieve specific sections.

License and pricing

The license is not stated (NOASSERTION). The project is free for personal use. Commercial use requires a paid one-time license.

Based on the project's public README and repository metadata.

Add this badge to your README

Featured on YunoDB Badge styles & help

Maintain this project? Show that it is featured on YunoDB.

Markdown[![Featured on YunoDB](https://yunodb.com/badge/jgravellejdocmunch-mcp.svg)](https://yunodb.com/tools/jgravellejdocmunch-mcp?utm_source=github&utm_medium=badge)
HTML<a href="https://yunodb.com/tools/jgravellejdocmunch-mcp?utm_source=github&utm_medium=badge"><img src="https://yunodb.com/badge/jgravellejdocmunch-mcp.svg" alt="Featured on YunoDB" height="20"></a>