Gemini MCP Server
About
# Gemini MCP Server This project implements a Model Context Protocol (MCP) server for Google Search enabling integration with AI assistants and other MCP-compatible clients. --- ## Prerequisites - Node.js 18+ - npm --- ## Installation ### 1. Clone the repository ```sh git clone…
Details
- Author
- rtbui2012
- Downloads
- 315
- Categories
- AI
Jump to
- Provides a googleSearchContent tool for web search results (title, link, snippet).
- Provides a googleSearchImages tool for image search (Markdown image link).
- Built on Node.js 18+ with npm package management.
- Works with any MCP-compatible client.
- Simple .env‑based credential management.
Setting up with Highlight
This MCP is not yet compatible with Highlight’s one-click setup. However, you can still use it with Highlight by following these steps:
- Download and install Highlight from highlightai.com/download
- Navigate to the plugins tab and select "Add Custom Plugin"
-
Configure the plugin with the settings below
Plugin Name
Gemini MCP ServerCommand (node, npx, python, etc.)Please refer to the README for specific instructions on how to obtain API keys or other required environment variables.
- Enable "Start Automatically" if you want the plugin to start when Highlight launches
From the repository
Install dependencies with npm install, set up your Google Custom Search API key in a .env file as GEMINI_API_KEY, then run a test search via npx ts-node src/test.ts or integrate it as an MCP server in your client configuration (e.g., claude_desktop_config.json).
Claude Desktop / Cursor
Paste into your MCP client config file to install this server.
{
"mcpServers": {
"gemini mcp server": {
"gemini-mcp-server-rtbui2012": {
"command": "npx",
"args": [
"ts-node",
"src/test.ts"
]
}
}
}
}McpServers
{
"gemini-mcp-server-rtbui2012": {
"command": "npx",
"args": [
"ts-node",
"src/test.ts"
]
}
}Gemini MCP Server
This project implements a Model Context Protocol (MCP) server for Google Search enabling integration with AI assistants and other MCP-compatible clients.
---
Prerequisites
- Node.js 18+
- npm
---
Installation
1. Clone the repository
git clone https://github.com/your-username/mcp-google-search.git
cd google-mcp-server
2. Install dependencies
npm install
3. Set up your Google Custom Search API credentials
Create a .env file in the project root with the following content:
GEMINI_API_KEY=your_google_api_key_here
- GOOGLE_API_KEY: Your Google Custom Search API key.
You can obtain these from the Google Custom Search JSON API documentation.
---
Usage
1. Run a test search
npx ts-node src/test.ts
This will execute a sample search and print the results to the console.
2. Integrate with MCP-compatible clients
To use this server as part of an MCP configuration, add an entry to your MCP settings file (e.g., claude_desktop_config.json):
{
"mcpServers": {
"gemini-api": {
"command": "npx",
"args": ["ts-node", "src/index.ts"],
"env": {
"GEMIN_API_KEY": "your_gemini_api_key_here",
}
}
}
}
Replace the API key and CX ID with your actual credentials.
---
Available Tools
The server provides the following tools:
1. googleSearchContent: Performs a web search using the Google Custom Search API and returns structured results (titles, links, and snippets).
2. googleSearchImages: Performs an image search using the Google Custom Search API and returns a Markdown image link for the first result.
Refer to the source code in src/tools.ts for detailed usage information.
---
Development
To make changes to the project:
1. Modify the code in the src directory as needed.
2. If you add or remove dependencies, update package.json accordingly.
3. Restart your MCP client or test script to apply changes.
---
Troubleshooting
- Ensure your Google API key and CX ID are correctly set in the .env file or your environment.
- Check that all dependencies are installed (npm install).
- Verify that you are using Node.js 18 or newer.
- If you make changes to the code, restart your server or test script.
---
License
This project is licensed under the MIT License.
Sign in to leave a review
Use Google, GitHub, or an email account so ratings stay tied to real people.
No reviews posted yet.
