home / mcp / gemini mcp tool windows fixed mcp server

Gemini MCP Tool Windows Fixed MCP Server

Gemini MCP Tool - Windows Fixed Version for Trae AI and Claude Desktop

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "orzcls-gemini-mcp-tool-windows-fixed": {
      "command": "npx",
      "args": [
        "-y",
        "[email protected]"
      ],
      "env": {
        "GEMINI_API_KEY": "YOUR_ACTUAL_API_KEY_HERE"
      }
    }
  }
}

You set up the Gemini MCP Tool Windows Fixed version as an MCP server to let AI assistants interact with Google’s Gemini CLI from Windows. This server bridges MCP clients with Gemini, handling command execution, file references, and sandbox-safe interactions while ensuring Windows compatibility and robust error handling.

How to use

Install and run the MCP server locally, then connect it to your MCP client (such as Trae AI or Claude Desktop). You can start by using NPX to run the fixed Windows version without installing it globally, or install it globally for ease of use. Once running, tell your MCP client to use the gemini-mcp-tool-windows-fixed endpoint to send prompts, file references, and sandboxed actions to Gemini.

How to install

Prerequisites you need before installing the MCP server on Windows:

Node.js v16.0.0 or higher
Google Gemini CLI installed and configured
A Gemini API key from Google AI Studio

# Check Node.js version
node --version  # Should be v16+

How to install

Option 1: Quick start with NPX (recommended) lets you run the fixed Windows MCP tool without a local install.

# Use latest version (recommended)
npx [email protected]

# Or use latest tag
npx -y gemini-mcp-tool-windows-fixed@latest

How to install

Option 2: Global installation installs the MCP tool so you can run it from any prompt.

# Install latest version globally
npm install -g [email protected]

# Run the tool
gemini-mcp-tool-windows-fixed

Additional setup and configuration

If you use the MCP with Claude Code or Claude Desktop, configure with one of the provided MCP server mappings. You will need your Gemini API key in the environment variable GEMINI_API_KEY.

# Example environment variable setup for a user session (PowerShell)
$env:GEMINI_API_KEY = "YOUR_ACTUAL_API_KEY_HERE"

Testing the installation

Verify that Gemini CLI and the MCP tool run correctly and that the MCP client can communicate with the server.

# Test Gemini CLI
gemini -p "Hello, how are you?"

# Test MCP server start (via npx form)
npx -y [email protected]

# Should show a message indicating the MCP server started

Troubleshooting

If you encounter common issues, verify your Node.js version, ensure the Gemini CLI is installed, and confirm your API key environment variable is set correctly.

# Check API key
echo $env:GEMINI_API_KEY

# Set if empty
$env:GEMINI_API_KEY = "your-api-key"

Windows-specific notes

This fixed version includes Windows-specific improvements such as better PowerShell parameter handling, UTF-8 encoding for Unicode text, quote escaping in command arguments, and enhanced environment variable loading.

Security and usage notes

Treat your Gemini API key as a secret. Do not expose it in logs or shared configurations. Use sandbox mode for risky operations and only enable it when you trust your workflows.

Notes on updates and compatibility

The Windows-fixed MCP tool maintains compatibility with various terminals (PowerShell, CMD, VS Code Terminal, etc.) and provides enhanced error handling and PATH management to simplify setup on Windows.

Available tools

ask-gemini

Interact with Google Gemini for analysis and questions.

brainstorm

Generate creative ideas using various brainstorming frameworks.

fetch-chunk

Retrieve cached chunks from changeMode responses.

timeout-test

Test timeout prevention mechanisms.

ping

Test connection to the server.

Help

Display help information about available tools.