This project provides an unofficial server for deepwiki.com, enabling users to fetch and process knowledge from the site into Markdown format for use in code editors like Cursor. It's designed for developers and researchers seeking to integrate up-to-date technical documentation into their workflows.

How It Works

The server crawls specified deepwiki.com URLs, sanitizes HTML content by removing non-essential elements like navigation and ads, and rewrites internal links for Markdown compatibility. It offers flexibility in output, allowing users to receive either a single aggregated Markdown document or structured data for individual pages, with configurable crawling depth and concurrency for performance.

Quick Start & Requirements

• Install/Run: npx mcp-deepwiki@latest or npm install -g mcp-deepwiki
• Prerequisites: Node.js, npm/npx.
• Configuration: Can be integrated with MCP clients via .cursor/mcp.json or run directly via HTTP.
• Docs: https://github.com/regenrek/mcp-deepwiki

Highlighted Details

• Domain safety restricts crawling to deepwiki.com.
• Supports "aggregate" (single document) and "pages" (structured data) output modes.
• Configurable via environment variables for concurrency, timeouts, and retries.
• Integrates as an MCP tool for seamless use with compatible clients.

Maintenance & Community

The project is maintained by regenrek. Links to X/Twitter and Bluesky are provided.

Licensing & Compatibility

MIT License. Permissive for commercial use and integration with closed-source projects.

Limitations & Caveats

The Docker deployment is noted as untested. Performance may vary based on repository size and network conditions, with timeouts and concurrency being adjustable parameters.