climaBot

About the project

climaBot is an experimental Python project for studying weather forecasting and its application to prediction markets. It combines meteorological data and probabilistic forecasts with market scanning, backtesting, strategy evaluation, settlement tracking and risk analysis.

The project follows a structured research process: establish a reproducible baseline, preregister hypotheses, test one improvement at a time and compare results using consistent metrics. It distinguishes weather predictions from market signals, simulated positions and settled outcomes so that each stage can be evaluated separately.

ClimaBot is currently limited to paper trading and the experimental evaluation of strategies through simulated positions. Its source code is hosted in a private GitHub repository; access can be requested from the author.