Sierra Leone Agri-ML: Code and Data Pipeline (v1.0.0)

First public release of the code and data pipeline for the paper "Can Machine Learning Forecast Rice Yields in Data-Constrained Settings? Satellite Climate Data, National Crop Statistics, and Lessons from Sierra Leone" (arXiv:2606.13959). Includes: FAOSTAT production data pipeline (2000–2024, nine major crops) CHIRPS rainfall and NASA POWER temperature integration XGBoost, Gradient Boosting, and Random Forest models with anti-leakage, expanding-window walk-forward evaluation Figures and results reproduction scripts This release is archived on Zenodo with a citable DOI.