Docmole

Dig through any documentation with AI


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

$3

* 🦙 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):

``bash
bunx docmole --help
`

Or install globally:

`bash
bun install -g docmole
`

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

Getting started

$3

Index and query any documentation site. Requires OPENAI_API_KEY.

`bash


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:

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

$3

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

`bash
bunx docmole -p agno-v2
`

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

CLI

Docmole has a built-in CLI for all operations:

`bash


Mintlify mode (proxy to Mintlify API)

docmole -p

Local RAG mode

docmole setup --url --id
docmole serve --project
docmole list
docmole stop --project
`

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 |

$3

`
~/.docmole/
├── projects/
│ └── /
│ ├── 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.