<p align="center"> <a href="https://github.com/Vigtu/docmole"> <img loading="lazy" alt="docmole" src="https://raw.githubusercontent.com/Vigtu/docmole/main/assets/docmole-hero.svg" width="100%"/> </a> </p>

Docmole

<p align="center"> <em>Dig through any documentation with AI</em> </p>

Docmole is an MCP server that lets you query any documentation site from AI assistants like Claude, Cursor, or any MCP-compatible client. The mole digs through docs so you don't have to.

Features

• 🔍 Universal docs support — works with any documentation site
• 🏠 Self-hosted RAG — LanceDB vectors + OpenAI embeddings, no Python needed
• ⚡ Zero-setup mode — instant access to Mintlify-powered sites
• 🧠 Multi-turn conversations — remembers context across questions
• 🔗 WebFetch compatible — links converted to absolute URLs
• 🔌 MCP native — works with Claude, Cursor, and any MCP client

Coming soon

• 🦙 Ollama support — fully local mode, no API keys needed
• 📄 Generic HTML extraction — support for non-Mintlify documentation sites
• 🔄 Incremental updates — only re-index changed pages

Installation

To use Docmole, run it directly with bunx (no install needed):

bunx docmole --help

Or install globally:

bun install -g docmole

Works on macOS, Linux and Windows. Requires Bun runtime.

Getting started

Local RAG Mode (any docs site)

Index and query any documentation site. Requires OPENAI_API_KEY.

# One-time setup — discovers pages and builds vector index
bunx docmole setup --url https://docs.example.com --id my-docs

# Start the MCP server
bunx docmole serve --project my-docs

Add to your MCP client:

{
  "mcpServers": {
    "my-docs": {
      "command": "bunx",
      "args": ["docmole", "serve", "--project", "my-docs"]
    }
  }
}

Mintlify Mode (zero setup)

For sites with Mintlify AI Assistant — no API key needed:

bunx docmole -p agno-v2

{
  "mcpServers": {
    "agno-docs": {
      "command": "bunx",
      "args": ["docmole", "-p", "agno-v2"]
    }
  }
}

CLI

Docmole has a built-in CLI for all operations:

# Mintlify mode (proxy to Mintlify API)
docmole -p <project-id>

# Local RAG mode
docmole setup --url <docs-url> --id <project-id>
docmole serve --project <project-id>
docmole list
docmole stop --project <project-id>

Run docmole --help for all options.

How it works

┌─────────────┐     ┌─────────────┐     ┌──────────────────────┐
│ MCP Client  │────▶│   Docmole   │────▶│ Embedded: LanceDB    │
│ (Claude,    │◀────│ MCP Server  │◀────│ Mintlify: API proxy  │
│  Cursor...) │     └─────────────┘     └──────────────────────┘
└─────────────┘

Local RAG Mode: Crawls documentation, generates embeddings with OpenAI, stores in LanceDB. Hybrid search combines semantic and keyword matching.

Mintlify Mode: Proxies requests to Mintlify's AI Assistant API. Zero setup, instant results.

Known Mintlify Project IDs

Documentation	Project ID
Agno	agno-v2
Resend	resend
Mintlify	mintlify
Vercel	vercel
Upstash	upstash
Plain	plain

Find more: Open DevTools → Network tab → use the AI assistant → look for leaves.mintlify.com/api/assistant/{project-id}/message

Configuration

Environment Variable	Default	Description
OPENAI_API_KEY	—	Required for local RAG mode
DOCMOLE_DATA_DIR	~/.docmole	Data directory for projects

Project structure

~/.docmole/
├── projects/
│   └── <project-id>/
│       ├── config.yaml      # Project configuration
│       └── lancedb/         # Vector database
└── global.yaml              # Global settings

Documentation

See AGENT.md for detailed documentation including:

• Architecture details
• Backend implementations
• Enterprise deployment guides

Contributing

PRs welcome! See the contributing guide for details.

Acknowledgments

• Mintlify for amazing documentation tooling
• Anthropic for Claude and the MCP protocol
• LanceDB for the vector database

License

The Docmole codebase is under MIT license.