Skip to main content

Contributing to PromptDrifter

This document provides guidelines for contributing to PromptDrifter.

Project Overview​

PromptDrifter is a tool for testing and monitoring LLM prompt consistency across different models and providers. It helps catch "prompt drift" - changes in model outputs for the same prompt over time.

How to Contribute​

  • Add support for new LLM providers/adapters
  • Improve documentation
  • Report bugs
  • Fix bugs or implement new features
  • Add tests
  • Share your experience using PromptDrifter

Contribution Process​

  1. Fork this repository
  2. Create an issue
  3. Make your changes
  4. Run the linter and tests
  5. Submit a pull request
  6. Address any feedback

Development Setup​

  1. Clone the repository:

    git clone https://github.com/your-username/promptdrifter.git
    cd promptdrifter

  2. Install dependencies using uv:

    pip install uv
    uv pip install -e .
    uv pip install -e ".[dev]"

Code Standards​

We use:

  • Ruff: For code formatting and linting
  • Pytest: For testing
  • Type Annotations: All functions must include type annotations
  • Pydantic: For data modeling and validation
  • Async/Await: For asynchronous programming

Commands:

make lint    # Run linter
make test-unit    # Run tests

Schema Versioning​

PromptDrifter uses a versioned schema system to ensure backward compatibility. For detailed information, see Schema Versioning Guide.

Working with Adapters​

When adding a new adapter:

  1. Create a new adapter class that inherits from Adapter base class
  2. Implement the required methods
  3. Add the adapter type to the schema
  4. Create unit tests for the adapter
  5. Add documentation and update README adapter table

Documentation Style​

  • Self-documenting code through clear naming
  • Minimal comments, only when necessary
  • README files for major components

Contact​

  • Open an issue or discussion on GitHub