# jDocmunch MCP

> jDocmunch MCP is an MCP server for developer tools. Index docs by heading and fetch just the section a coding agent needs. It runs locally on your machine and is started by your AI client.

- **Kind:** MCP server
- **Pricing:** Free trial
- **Platforms:** CLI, self-hosted
- **Repository:** https://github.com/jgravelle/jdocmunch-mcp
- **MCP Registry name:** io.github.jgravelle/jdocmunch-mcp
- **Categories:** [developer tools](https://ai-powered.com/categories/developer-tools), [knowledge and memory](https://ai-powered.com/categories/knowledge-and-memory)
- **Listing updated:** 2026-11-15

## About

Most agents burn context by opening whole documentation files. jDocMunch parses a doc set into a tree keyed by heading hierarchy, stores each section's byte offsets, and returns only the section, subtree or ranked hit requested, read directly from the original file. Section IDs stay stable across re-indexing as long as path, heading text and level are unchanged.

It indexes Markdown, reStructuredText, AsciiDoc, Org, notebooks, HTML, YAML, JSON and OpenAPI specs. The README publishes four benchmarks (Kubernetes docs, SciPy, LangChain MDX and a wiki) and says savings depend on how large a file is relative to the section needed; one Kubernetes query saved 27,285 tokens. It also states each release must clear an nDCG 0.95 retrieval gate.

Install with `uv tool install jdocmunch-mcp` and run `jdocmunch-mcp init`, which detects clients and writes their config. Folders are capped at 10,000 files unless raised. Personal use is free; commercial use needs a paid licence.

## Features

- **Section-level retrieval:** Returns one heading section or subtree instead of a whole file.
- **Byte-precise extraction:** Reads content from exact offsets in the source file.
- **Stable section IDs:** IDs persist across re-indexing when path, heading text and level are unchanged.
- **Many doc formats:** Handles Markdown, rst, AsciiDoc, Org, notebooks, HTML, YAML, JSON and OpenAPI.
- **Local and repo indexing:** Index a local folder or a GitHub repository.
- **Compact search rows:** A compact mode shrinks search results to save tokens.

## Good for

- Fetch the authentication section from a large docs set
- Browse a project's structure by heading tree
- Cut token use when an agent reads Kubernetes-scale docs
- Index a GitHub repo's documentation for an assistant
- Search OpenAPI specs by section

## Install

### Claude Code (Terminal)

```sh
claude mcp add jdocmunch-mcp -- uvx jdocmunch-mcp
```

### Claude Desktop (claude_desktop_config.json)

```json
{
  "mcpServers": {
    "jdocmunch-mcp": {
      "command": "uvx",
      "args": [
        "jdocmunch-mcp"
      ]
    }
  }
}
```

### Cursor (~/.cursor/mcp.json)

```json
{
  "mcpServers": {
    "jdocmunch-mcp": {
      "command": "uvx",
      "args": [
        "jdocmunch-mcp"
      ]
    }
  }
}
```

### VS Code (.vscode/mcp.json)

```json
{
  "servers": {
    "jdocmunch-mcp": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "jdocmunch-mcp"
      ]
    }
  }
}
```

### Windsurf (~/.codeium/windsurf/mcp_config.json)

```json
{
  "mcpServers": {
    "jdocmunch-mcp": {
      "command": "uvx",
      "args": [
        "jdocmunch-mcp"
      ]
    }
  }
}
```

### Other clients (your client's MCP settings (most use this shape))

```json
{
  "mcpServers": {
    "jdocmunch-mcp": {
      "command": "uvx",
      "args": [
        "jdocmunch-mcp"
      ]
    }
  }
}
```

## Tools

- `search_sections`: Ranked search over indexed documentation sections.
- `get_section`: Fetch one section's full content from its byte offsets.
- `get_toc_tree`: Structural view of a whole documentation corpus.
- `index_repo`: Index documentation from a GitHub repository.
- `index_local`: Index a local documentation folder.
- `jdocmunch_guide`: Returns instructions that steer the agent to use the doc tools.

## FAQ

### Is it free?

Free for personal use; the README says commercial use requires a paid licence, with terms in the repository.

### What does it need?

Python 3.10 or later and any MCP-compatible client.

### Is there a file limit?

A folder is capped at 10,000 files, newest first, and the cap can be raised with --max-files.

### How do I index a very large folder?

Run jdocmunch-mcp index-local from a shell to avoid client tool timeouts.

### Is jDocmunch MCP free?

jDocmunch MCP is a paid product with a free trial.

### How do I add jDocmunch MCP to Claude Code?

Run this command in your terminal: claude mcp add jdocmunch-mcp -- uvx jdocmunch-mcp

### What are the best alternatives to jDocmunch MCP?

The closest alternatives to jDocmunch MCP are REA, Context7, Chrome DevTools MCP, Codebase Memory and GitHub.

## Where this comes from

- Description: written from the project's README.
- Install steps: from the official MCP Registry entry.

## Alternatives

- [REA](https://ai-powered.com/mcp/rea): REA is an open-source MCP server for developer tools.
- [Context7](https://ai-powered.com/mcp/context7): Context7 is a freemium MCP server for developer tools.
- [Chrome DevTools MCP](https://ai-powered.com/mcp/chrome-devtools-mcp-2): Chrome DevTools MCP is an open-source MCP server for browser automation.
- [Codebase Memory](https://ai-powered.com/mcp/codebase-memory): Codebase Memory is an open-source MCP server for knowledge and memory.
- [GitHub](https://ai-powered.com/mcp/github-2): GitHub is an open-source MCP server for developer tools.
- [Serena MCP: the IDE for your agent](https://ai-powered.com/mcp/serena-mcp-the-ide-for-your-agent): Serena MCP: the IDE for your agent is an open-source MCP server for developer tools.
- [awesome-mcp-servers](https://ai-powered.com/mcp/awesome-mcp-servers-2): awesome-mcp-servers is an open-source MCP server for developer tools.
- [XERJ](https://ai-powered.com/mcp/xerj): XERJ is an open-source MCP server for knowledge and memory.

---
Source: https://ai-powered.com/mcp/jdocmunch-mcp
