google_search_mcp_server

google_search_mcp_server

1
research_and_data

The Google Search MCP Server is a tool that facilitates real-time web and image searches through Google's Custom Search API. It enhances AI assistants by providing direct search capabilities, enabling them to fetch the latest information from the web efficiently. The project integrates seamlessly with tools like Claude to expand an assistant's functionality.

Google Search MCP Server

A Model Context Protocol (MCP) server providing web and image search capabilities via Google's Custom Search API, enabling AI assistants to access current web information.

Features

  • google_web_search: Search the web for up-to-date information.
  • google_image_search: Search for images related to queries.

Usage

Connect this server to an MCP-compatible client like Claude on VSCode to enable these search functionalities.

Prerequisites

  • Node.js (v18 or higher), npm, Google Cloud Platform account, Custom Search API key.

Setup

  1. Clone the repository and install dependencies.
  2. Configure API credentials via .env file.

Follow detailed setup guides to integrate the server with different clients such as Claude for Desktop, VSCode, and Cursor. Testing instructions provided for both automated and manual testing scenarios.

Troubleshooting

Includes guidance on common issues like environment variables, API errors, and connection issues.

Limitations

  • Google Custom Search API free tier limits to 100 queries per day.
  • Server rate limit of 5 requests per second.