Summary

Gemini-MCP integrates Google's Gemini CLI with Claude Code via the Model Context Protocol (MCP). It enables Claude Code to invoke Gemini for AI-assisted programming, particularly leveraging Gemini's strengths in frontend design and code generation. This collaboration aims to enhance development efficiency and output quality by providing a structured interface for AI-to-AI interaction.

How It Works

Gemini-MCP acts as an MCP server, wrapping the Gemini CLI for programmatic access by Claude Code. This architecture leverages Gemini's proficiency in frontend aesthetics, task planning, and requirement clarification, complementing Claude Code's workflow. The system supports session management and structured results, facilitating complex AI-driven development pipelines.

Quick Start & Requirements

• Prerequisites: Claude Code, Google Gemini CLI, and the uv tool installed. uv installation varies by OS (PowerShell for Windows, shell script for Linux/macOS). WSL is recommended for Windows.
• Installation: Run claude mcp add gemini -s user --transport stdio --uvx --from git+https://github.com/GuDaStudio/geminimcp.git geminimcp.
• Verification: Use claude mcp list.
• Optional Configuration: Add mcp__gemini__gemini to ~/.claude/settings.json for auto-approval.
• Recommended Prompt: Enhance Claude Code's collaboration by adding a detailed prompt to ~/.claude/CLAUDE.md, guiding Gemini usage for requirement clarification, planning, and frontend prototyping.
• Links: Implicitly via GitHub repository.

Highlighted Details

• Demonstrates Gemini's advanced frontend design capabilities, significantly improving web page aesthetics and animations in an example.
• Encapsulates Gemini CLI into a standard MCP protocol for programmatic invocation and structured data exchange with Claude Code.
• Supports session management (SESSION_ID) for multi-turn conversations.
• Includes a detailed prompt configuration to optimize Claude Code's collaboration with Gemini for specific tasks.

Maintenance & Community

No specific details on maintainers, sponsorships, or community channels are provided in the README. Contribution involves cloning the repository and using uv sync for dependencies.

Licensing & Compatibility

Licensed under the MIT License, generally permitting commercial use and modification.

Limitations & Caveats

Gemini has an effective context length limitation (32k). The project advises against using Gemini for complex backend business logic, focusing its strengths on frontend tasks, planning, and requirement clarification.