gemini.md to teach Claude to use google's gemini cli as his tool

gemini.md

Gemini CLI for Large Codebase Analysis

Use gemini -p for analyzing large codebases that exceed Claude's context limits.

File/Directory Inclusion

Use @ syntax with paths relative to your working directory:

# Single file
gemini -p "@src/main.py Explain this file's purpose"

# Multiple files  
gemini -p "@package.json @src/index.js Analyze dependencies"

# Directories
gemini -p "@src/ Summarize the architecture"

# Entire project
gemini -p "@./ Project overview"
# or: gemini --all_files -p "Analyze project structure"
Feature Verification Examples
bash# Check implementations
gemini -p "@src/ Is dark mode implemented? Show relevant files"
gemini -p "@api/ Is JWT auth implemented? List endpoints"
gemini -p "@src/ Any React WebSocket hooks? Show with paths"
gemini -p "@backend/ Is rate limiting implemented?"

# Security checks
gemini -p "@api/ Are SQL injection protections implemented?"
gemini -p "@src/ Show error handling patterns"

# Test coverage
gemini -p "@src/payment/ @tests/ Is payment module tested?"
When to Use

Analyzing entire codebases or large directories
Files totaling >100KB
Need project-wide pattern analysis
Verifying feature implementations across codebase
Current context insufficient

Note: Paths are relative to your current directory when running gemini.