Weather NO-Side Temperature Arbitrage

This bot targets Polymarket temperature prediction markets, specifically those that are priced too high for a given temperature outcome. It uses live METAR (aviation weather) observations and Open-Meteo forecast data to build a Gaussian probability model of the actual temperature. When the market price implies a probability higher than the bot's modeled probability, it places a NO bet (i.e., the event will NOT happen). Entry is triggered when the mispricing exceeds a configurable threshold. Exit occurs at settlement or if the bot dynamically adjusts its position as new data arrives. The strategy is purely paper-trading (simulated) and is written in Python. The GitHub repository includes a full pipeline for fetching weather data, computing probabilities, and executing mock orders. No live profit/loss data is attached, but the methodology is statistically grounded.