MCP Web Search

A Model Context Protocol (MCP) server for web search using DuckDuckGo. No API key required.

Features

• Web Search - Search the web via DuckDuckGo's static HTML interface

• Content Extraction - Fetch and extract readable content from any webpage using Mozilla's Readability

• Zero Configuration - No API keys, no signup, works out of the box

• Lightweight - Uses Node.js native fetch, no heavy browser automation

Related MCP server: Web Scout MCP Server

Tools

Usage

Local installation

git clone https://github.com/doranobik/mcp-web-search.git
cd mcp-web-search
npm install
npm run build
node dist/index.js

Note: This package is not published to npm. The mcp-web-search name on npm belongs to an unrelated project, so install from this repository instead.

Docker

docker build -t mcp-web-search .
docker run -i mcp-web-search

MCP Client Configuration

Claude Desktop (claude_desktop_config.json):

{
  "mcpServers": {
    "web-search": {
      "command": "node",
      "args": ["/path/to/mcp-web-search/dist/index.js"]
    }
  }
}

LM Studio (mcp.json):

{
  "mcpServers": {
    "web-search": {
      "command": "node",
      "args": ["/path/to/mcp-web-search/dist/index.js"]
    }
  }
}

Tool Parameters

webSearch

fetchWebContent

Disclaimer

This server searches the web by parsing DuckDuckGo's HTML interface, not an official API. It may break if DuckDuckGo changes its markup, and heavy automated use may violate DuckDuckGo's terms of service or lead to rate limiting. Use it responsibly and at your own risk.

License

MIT — see LICENSE for details.