georgedouzas/sports-betting ? reverse-engineered prompt

Build me a Python toolkit for sports betting that can pull in match stats and odds from chosen sources, shape the data for modeling, and backtest betting strategies.

I want to be able to use it from Python and also from an AI agent or command line. The main idea is that I can load historical data, train a betting model, see which bets look like value bets for upcoming games, and then run a backtest to understand how the strategy would have performed. It should work with normal scikit learn style models so I can plug in things like logistic regression or other classifiers.

Please include a simple way to define data sources, combine stats and odds, split train and upcoming fixtures, and evaluate the results with clear cash and stake based outputs. If it makes sense, add a small agent friendly interface for common tasks and keep the code clean, well tested, and documented. Look up current docs online if you need to.