Description

Docs MCP Server provides a specialized documentation scraping and retrieval system that enables AI assistants to access library documentation from various sources including GitHub, NPM, PyPI, and web pages. Built with TypeScript, it implements a pipeline architecture that handles document scraping, processing, splitting, and storage with features like semantic markdown splitting, greedy chunk optimization, and version-aware retrieval. The server exposes tools for searching documentation, finding specific versions, listing available libraries, and managing scraping jobs through a well-defined API. Particularly valuable for developers who need AI assistants to reference accurate, up-to-date documentation without leaving their workflow.

Installation

npx -y @arabold/docs-mcp-server

npm install @arabold/docs-mcp-server

git clone https://github.com/arabold/docs-mcp-server

Claude Desktop Configuration

Add this configuration to your Claude Desktop config file to enable this MCP server:

Config file location: ~/Library/Application Support/Claude/claude_desktop_config.json

Config file location: %APPDATA%\Claude\claude_desktop_config.json

Config file location: ~/.config/Claude/claude_desktop_config.json

claude_desktop_config.json

{
  "mcpServers": {
    "documentation-scraper": {
      "command": "npx",
      "args": [
        "-y",
        "@arabold/docs-mcp-server"
      ]
    }
  }
}