WebSearch
[Self-hosted] A Model Context Protocol (MCP) server implementation that provides a web search capability over stdio transport. This server integrates with a WebSearch Crawler API to retrieve search results.
claude mcp add --transport stdio mnhlt-websearch-mcp npx websearch-mcp
How to use
WebSearch-MCP exposes a web search capability over the MCP stdio transport. It acts as a bridge between AI assistants that support MCP and a WebSearch crawler API. The server provides a single MCP tool named web_search that accepts a query and optional parameters to fetch real-time web results. By configuring API_URL, you direct the MCP server to the crawler service, and MAX_SEARCH_RESULT controls how many results are returned when the request does not specify a limit. Clients such as Claude Desktop, Cursor IDE, or other MCP-enabled tools can invoke the web_search tool to retrieve up-to-date information from the web within your conversations.
How to install
Prerequisites:
- Node.js and npm installed on your machine
- Optional: Smithery CLI for automated installation
Install via Smithery (recommended for Claude Desktop users):
npx -y @smithery/cli install @mnhlt/WebSearch-MCP --client claude
Manual installation (global npm install):
npm install -g websearch-mcp
Run directly with npx (no permanent install):
npx websearch-mcp
If you prefer to run locally after installing npm package globally, ensure the crawler service is reachable at the API_URL you configure (default http://localhost:3001).
Additional notes
Tips:
- The MCP server uses the environment variable API_URL to reach the WebSearch crawler API. Set it to the crawler service URL (default http://localhost:3001).
- MAX_SEARCH_RESULT controls the default number of results when a request does not specify numResults. Adjust to balance result depth and token usage.
- To run the full setup, you must also start the Docker-based crawler service as documented in the README, or point API_URL to an already running crawler instance.
- If you encounter Windows commandline issues, you may need to use the Windows-friendly MCP configuration shown in the README (cmd /c npx websearch-mcp).
- Health and testing: verify the crawler API health with http://localhost:3001/health and test searches with the curl example provided in the setup guide.
- This MCP server uses stdio transport; ensure your MCP client is configured to communicate over stdin/stdout.
Related MCP Servers
zen
Selfhosted notes app. Single golang binary, notes stored as markdown within SQLite, full-text search, very low resource usage
MCPBench
The evaluation benchmark on MCP servers
MCP -Deepseek_R1
A Model Context Protocol (MCP) server implementation connecting Claude Desktop with DeepSeek's language models (R1/V3)
mcp-fhir
A Model Context Protocol implementation for FHIR
mcp
Inkdrop Model Context Protocol Server
mcp-appium-gestures
This is a Model Context Protocol (MCP) server providing resources and tools for Appium mobile gestures using Actions API..