CONTRIBUTING

# Contributing to Kernel ML

Thank you for your interest in Kernel ML. We are a collective focused on bridging the gap between machine learning research and production-ready implementation. 

We value code that is as readable as the math behind it.

## Our Philosophy

* **Clarity over Complexity:** We prefer simple, well-documented code over "clever" one-liners.
* **Single Stack Focus:** We build primarily with Python, PyTorch, and FastAPI.
* **Rigorous Testing:** Since we bridge the gap to production, we prioritize unit tests and type hinting.

## How to Contribute

1.  **Find an Issue:** Browse our open issues or start a discussion if you have a new idea.
2.  **Fork & Branch:** Create a feature branch from `main`.
3.  **Code Standards:** * Use `black` for formatting.
    * Include docstrings (Google or NumPy style) for all new functions.
    * Add type hints to maintain codebase health.
4.  **Testing:** Ensure all tests pass. If you're adding a feature, please add a corresponding test in the `/tests` directory.
5.  **Submit a PR:** Provide a clear description of your changes and why they are necessary.

## Development Setup

We recommend using `poetry` or `uv` for dependency management:

```bash
# Example setup
git clone [https://github.com/kernel-ml/](https://github.com/kernel-ml/)[repo-name]
cd [repo-name]
pip install -e .[dev]