gemini

Execute Gemini CLI for AI-powered code analysis and generation with fixed 2-hour timeout.

• Run Gemini CLI for AI-powered code analysis and generation
• Integrates Google's Gemini models into Claude Code workflows
• Supports complex reasoning and code tasks with advanced AI
• Uses Google's latest AI technology for alternative code perspectives

• Fixed 7200000ms (2-hour) timeout, immutable
• Supports multiple Gemini models via GEMINI_MODEL env var
• Cross-platform (Windows/macOS/Linux) compatibility
• No external dependencies (uses standard library)
• Output streamed directly from Gemini CLI
• Mandatory uv run for environment management
• Alternative execution via direct script or python3
• PEP 723 compliant with inline metadata

1. 1.Run via uv with fixed timeout: uv run script.py '<prompt>' [working_dir]
2. 2.Set GEMINI_MODEL environment variable to select model variant
3. 3.Include timeout: 7200000 in bash tool calls for protection
4. 4.Use working_dir parameter to specify project directory
5. 5.Ensure Gemini CLI is installed and authenticated