AI Code Review Agent CLI is a tool that leverages Google Gemini AI to automatically analyze Git changes and provide professional review reports for code quality improvement.
This project aims to streamline the code review process by proactively identifying potential bugs, security vulnerabilities, and performance issues in developer code, and providing concrete improvement suggestions in your preferred language (default: Korean).
- Language: Go 1.25.4
- CLI Framework: Cobra
- AI Model: Google Gemini (Configurable, default:
gemini-2.5-flash) (via Google GenAI SDK) - Configuration: JSON based local config
- Go: Version 1.25 or higher (Download)
- Git: Required for project version control and tracking changes.
- Google AI API Key: API key required for using the Gemini model. (Get API Key)
You can install code-reviewer using Homebrew:
brew install Raonshi/tap/code-reviewerClone the project and install dependencies.
# Clone the repository
git clone https://github.com/your-username/code-reviewer.git
cd code-reviewer
# Download dependencies
go mod downloadBuild and run the project.
# Build the binary
go build -o code-reviewer main.go
# Run the help command to verify installation
./code-reviewer --helpOn the first run, you will be prompted to enter your Google AI API Key, select an AI Model (default: gemini-2.5-flash), and choose an Output Language (default: Korean).
# Run code review (default: all changes)
./code-reviewer report
# Run code review on staged changes
./code-reviewer report --staged
# Run code review on unstaged changes
./code-reviewer report --unstaged
# Generate technical documentation for code changes
./code-reviewer document
# Generate auto-fixes for code issues (Prints proposed fix)
./code-reviewer fix
# Manage configuration
./code-reviewer config list
./code-reviewer config get output_language
./code-reviewer config set output_language English- AI-Powered Code Review: Uses Google Gemini models (configurable) to deeply analyze code changes (
git diff). - Multi-Language Support: Supports output in various languages (configurable, default: Korean).
- Configuration Management: Easily manage API keys, models, and output languages via the
configcommand. - Automatic Grading: Changes are automatically graded into 4 levels: Good, Not Bad, Bad, Need Check to quickly identify importance.
- Concrete Improvement Suggestions: Beyond simple pointers, it provides immediately applicable code snippets and refactoring guides for areas needing correction.
- Auto-Fix Generation: Generates corrected code snippets for identified issues (currently prints to stdout).
- Documentation Generation: Automatically generates technical documentation for code changes using the
documentcommand. - Git Integration: Automatically detects Staged and Unstaged changes in the current Git repository.