An MCP server that provides AI assistants with direct access to technical documentation from sites like React, Python, and crawl4ai through intelligent web scraping and caching.

Installation

uvx (recommended)

uvx mcp-server-docy

pip

pip install mcp-server-docy
DOCY_DOCUMENTATION_URLS="https://docs.crawl4ai.com/,https://react.dev/" python -m mcp_server_docy

Docker

docker pull oborchers/mcp-server-docy:latest
docker run -i --rm -e DOCY_DOCUMENTATION_URLS="https://docs.crawl4ai.com/,https://react.dev/" oborchers/mcp-server-docy

Configuration

Claude Desktop - uvx

"mcpServers": {
  "docy": {
    "command": "uvx",
    "args": ["mcp-server-docy"],
    "env": {
      "DOCY_DOCUMENTATION_URLS": "https://docs.crawl4ai.com/,https://react.dev/"
    }
  }
}

Claude Desktop - Docker

"mcpServers": {
  "docy": {
    "command": "docker",
    "args": ["run", "-i", "--rm", "oborchers/mcp-server-docy:latest"],
    "env": {
      "DOCY_DOCUMENTATION_URLS": "https://docs.crawl4ai.com/,https://react.dev/"
    }
  }
}

VS Code - uvx

{
  "mcp": {
    "servers": {
      "docy": {
        "command": "uvx",
        "args": ["mcp-server-docy"],
        "env": {
          "DOCY_DOCUMENTATION_URLS": "https://docs.crawl4ai.com/,https://react.dev/"
        }
      }
    }
  }
}

Available Tools

Tool	Description
list_documentation_sources_tool	List all available documentation sources
fetch_documentation_page	Fetch documentation page content by URL in markdown format
fetch_document_links	Fetch all links from a documentation page

Features

• Instant documentation access - Direct access to React, Python, crawl4ai, and any other technology documentation
• Hot reload support - Add new documentation sources on-the-fly without restarting via .docy.urls file
• Intelligent caching - Reduces latency and external requests while keeping content up-to-date
• Self-hosted management - Keep documentation access within your security perimeter
• Seamless MCP integration - Works with Claude, VS Code, and other MCP-enabled tools
• Multiple transport protocols - Supports stdio and SSE (Server-Sent Events) protocols
• Persistent disk caching - Cache persists after server restart for improved performance

Environment Variables

Optional

• DOCY_DOCUMENTATION_URLS - Comma-separated list of documentation site URLs to include
• DOCY_DOCUMENTATION_URLS_FILE - Path to file containing documentation URLs, one per line (default: ".docy.urls")
• DOCY_CACHE_TTL - Cache time-to-live in seconds (default: 432000)
• DOCY_CACHE_DIRECTORY - Path to cache directory (default: ".docy.cache")
• DOCY_USER_AGENT - Custom User-Agent string for HTTP requests
• DOCY_DEBUG - Enable debug logging ("true", "1", "yes", or "y")
• DOCY_SKIP_CRAWL4AI_SETUP - Skip running crawl4ai-setup command on startup ("true", "1", "yes", or "y")
• DOCY_TRANSPORT - Transport protocol to use (options: "sse" or "stdio", default: "stdio")

Usage Examples

Are we implementing Crawl4Ai scraping results correctly? Let's check the documentation.

What does the documentation say about using mcp.tool? Show me examples from the documentation.

How should we structure data according to React documentation? What are the best practices?

Please use Docy to search...

Notes

Claude may default to using the built-in WebFetchTool instead of Docy. To explicitly request Docy functionality, use a phrase like "Please use Docy to search...". The server supports hot-reloading of URL configuration files, allowing you to add/modify documentation sources without restarting. Global server configuration is available for commands via server/README.md for running a persistent SSE server.