Gemini Bridge

Bridges Claude with Google's Gemini AI through the official Gemini CLI, enabling direct queries and file-based context sharing between the two language models.

This lightweight MCP server by Shelakh provides a direct bridge between Claude and Google's Gemini AI through the official Gemini CLI, offering two core tools for sending queries with or without file attachments. Built with Python and FastMCP, it executes gemini CLI commands via subprocess with configurable timeouts, normalizes model names (supporting both flash and pro variants), and concatenates file contents directly into prompts for context-aware queries. The implementation prioritizes simplicity over complexity with no caching, sessions, or persistent state, making it ideal for developers who want straightforward access to Gemini's capabilities from within Claude workflows without additional infrastructure dependencies.