Gemini MCP Server

A Model Context Protocol (MCP) server that provides access to Google's Gemini 3.5 Flash API. This server runs locally via npx for seamless integration with Claude Code.

Features

• Full MCP Protocol Support: JSON-RPC 2.0 compliant MCP server implementation
• Gemini 3.5 Flash: Access to Google's latest Gemini model
• Local Execution: Runs via npx for easy local development
• TypeScript Implementation: Fully typed codebase with robust error handling
• Configurable Parameters: Control temperature and max tokens

Setup

1. Install dependencies:

npm install

2. Create your configuration file:

cp config.json.example config.json

3. Edit config.json with your Gemini API key:

{
  "geminiApiKey": "your-gemini-api-key-here"
}

Get your API key from: https://aistudio.google.com/apikey

Local Development

Run the MCP server locally:

npm run dev

Build the TypeScript code:

npm run build

Usage with Claude Code

Add to your ~/.claude.json MCP servers configuration:

{
  "mcpServers": {
    "gemini": {
      "type": "stdio",
      "command": "cmd",
      "args": [
        "/c",
        "npx",
        "tsx",
        "C:/Users/mesol/workspace/gemini-mcp/src/index.ts"
      ],
      "env": {}
    }
  }
}

After adding the configuration, restart Claude Code to load the MCP server.

Available Tools

query_gemini

Query Google's Gemini 3.5 Flash API:

Parameters:

• prompt (required): The prompt to send to Gemini
• max_tokens (optional): Maximum tokens in response (default: 8192)
• temperature (optional): Temperature for response generation, 0.0 to 2.0 (default: 1.0)

Example Usage:

{
  "name": "query_gemini",
  "arguments": {
    "prompt": "Explain quantum computing",
    "max_tokens": 4096,
    "temperature": 0.7
  }
}

Architecture

• src/index.ts: Main MCP server entry point with stdio transport
• src/config.ts: Configuration loading utility
• config.json: API key configuration (gitignored)
• tsconfig.json: TypeScript configuration

Troubleshooting

Connection Issues

1. API Key Error: Verify your API key in config.json is correct
2. Config Not Found: Ensure config.json exists (copy from config.json.example)
3. Node Version: Requires Node.js >= 22.21.0

Getting Help

For issues with:

• Gemini API: Visit https://ai.google.dev/gemini-api/docs
• MCP Protocol: Visit https://modelcontextprotocol.io
• This server: Check the error logs in Claude Code

Files

• config.json.example: Template for configuration
• config.json: Your actual config (create from example, not tracked in git)
• .gitignore: Ensures config.json is not committed

License

MIT