Ambrogio

Ambrogio - Your Intelligent Dev Agent for Technical Debt

Ambrogio is an innovative development agent designed to assist teams in managing technical debt effectively. By implementing automated analysis, Ambrogio improves code quality and maintainability with an emphasis on enhancing docstring documentation.

🎯 Mission

Ambrogio's mission is to empower development teams to sustain high-quality codebases. It automatically identifies and addresses technical debt, beginning with the generation of comprehensive and precise docstrings. Leveraging advanced language models, Ambrogio comprehends the intricacies of your code to produce meaningful documentation.

✨ Features

Current Capabilities

• Smart Docstring Generation: Automatically generates clear and comprehensive docstrings for classes and methods.
• Multi-Provider Support: Compatible with various LLM providers through LiteLLM integration.
• Unit Test Generation (Beta): Examines code coverage and generates any required unit tests.

🚀 Future Enhancements

• Conducts pre-PR test runs to avert regressions.
• Enhances existing docstrings for modified methods.
• Generates thorough documentation for better comprehension.
• Facilitates type safety through refactoring.
• Cleans up spaghetti code.
• Enforces adherence to code formatting and best practices.

🚀 Usage Examples

Basic Execution

# Generate docstrings with default settings (utilizes OPENAI_API_KEY from environment)
ambrogio

# Execute with custom configuration
ambrogio \
  --path /path/to/your/project \
  --api-key your-api-key \
  --model gpt-4 \
  --max-api-calls 20

# Generate unit tests (Beta)
ambrogio \
  --mode coverage \
  --max-iterations 5

⚠️ Note: The unit test generation feature is in beta and should be run in a virtual environment with pytest installed.

Advanced Configuration

# Utilizing Azure OpenAI
ambrogio \
  --api-key your-azure-key \
  --api-base https://your-azure-deployment.openai.azure.com \
  --model gpt-4

# Utilizing Anthropic's Claude
ambrogio \
  --api-key your-anthropic-key \
  --model claude-2

Common Options

--path           Specify the path to the Python project (default: current directory)
--api-key        Provide the API key for the LLM provider (default: OPENAI_API_KEY from env)
--model          Specify the model to utilize (default: gpt-4o-mini)
--api-base       Provide the base URL for API endpoint (mandatory for Azure)
--mode           Choose operation mode ('docstring' or 'coverage'), default is 'docstring'

🔧 Supported LLM Providers

Ambrogio uses LiteLLM for versatile provider support, compatible with:

• OpenAI (GPT-3.5, GPT-4)
• Azure OpenAI
• Anthropic (Claude)
• And many more! For a complete list, visit the LiteLLM's documentation.

🤝 Contributing

Contributions are encouraged. Possible ways to contribute include:

• Reporting bugs
• Suggesting new features
• Submitting pull requests

By employing Ambrogio, development teams can enhance their coding efficiency and drastically reduce technical debt, paving the way for sustainable software development.