gemini-mcp-tool

Overview

This Model Context Protocol server bridges AI assistants with Google Gemini's powerful CLI, leveraging Gemini's extensive token window for analyzing large files and codebases. It provides tools for asking natural language questions, conducting analysis with file references using the @ syntax, and safely testing code in a sandbox environment. The server supports both tool calls for AI assistants and slash commands for direct user interaction within Claude Code.

Try asking AI

After installing, here are 5 things you can ask your AI assistant:

• ask-geminiprompt (required), model (optional), sandbox (optional)

  Asks Google Gemini for its perspective, with support for file references and general questions.

  Try:analyze @src/main.js and explain what it doesTry:use gemini to search for the latest tech news
• sandbox-testprompt (required), model (optional)

  Safely executes code or commands in Gemini's sandbox environment.

  Try:create and run a Python script that processes dataTry:test @script.py safely and explain its output
• Ping

  A simple test tool that echoes back a message.

  Try:pingTry:test connection
• Help

  Shows the Gemini CLI help text.

  Try:show helpTry:get gemini CLI documentation

Comparable tools

Installation

claude mcp add gemini-cli -- npx -y gemini-mcp-tool

{
  "mcpServers": {
    "gemini-cli": {
      "command": "npx",
      "args": ["-y", "gemini-mcp-tool"]
    }
  }
}

FAQ

What is the purpose of this MCP server?

It enables AI assistants to interact with Google Gemini CLI, leveraging Gemini's massive token window for large file analysis and codebase understanding.

How do I use file references with Gemini?

Use the @ syntax in your prompts to reference files or directories, such as 'analyze @src/main.js' or 'summarize @.' for the current directory.

Compare gemini-mcp-tool with