Thanks for your interest in improving ModelSentinel! 🎉
git clone https://github.com/sowaiba/modelsentinel.git
cd modelsentinel
python -m venv .venv && source .venv/bin/activate # Windows: .venv\Scripts\activate
pip install -e ".[dev]"- Tests pass:
pytest - Lint is clean:
ruff check . - New behaviour is covered by tests in
tests/. - Public functions have docstrings and type hints.
- Keep dependencies minimal (numpy / pandas / scikit-learn / scipy only, unless discussed in an issue first).
- One logical change per pull request.
- Every scoring function returns a bounded value in
[0, 100]for consistency. - Open an issue to discuss large features before building them.
Open an issue with a minimal reproducible example, your Python version, and the full traceback.