Add gemini-bridge MCP server

[eLyiN]

Summary

Add gemini-bridge MCP server - a lightweight bridge connecting AI coding assistants to Google's Gemini AI via the official CLI.

Features

• Zero API costs - Uses free Gemini CLI instead of paid API
• Stateless operation - No sessions, caching, or complex state management
• Universal compatibility - Works with Claude Code, Cursor, VS Code, and any MCP-compatible client
• Production ready - Robust error handling with 60-second timeouts
• Minimal dependencies - Only requires mcp>=1.0.0 and Gemini CLI
• Two core MCP tools:
  • consult_gemini - Direct queries to Gemini
  • consult_gemini_with_files - File-aware analysis and code review

Repository

• GitHub: https://github.com/shelakh/gemini-bridge
• PyPI: https://pypi.org/project/gemini-bridge/
• License: MIT
• Category: AI tools

Docker Implementation

• Multi-stage build with Node.js 20 + Python 3.11
• Installs Gemini CLI via npm globally
• Installs gemini-bridge from PyPI
• Proper user permissions and security configuration
• Entry point configured for MCP stdio transport

Test Plan

• Docker image builds successfully
• Gemini CLI properly installed and accessible
• Python environment configured correctly
• MCP server responds to requests
• Error handling works as expected
• No secrets or environment variables required

Ready for Docker team review and inclusion in the official MCP catalog.