Bot de trading météo Polymarket avec Claude

Write me a Python bot that trades Polymarket weather markets. Logic: 1. Every hour, pull the list of active temperature markets from Polymarket's CLOB API. 2. For each market, identify the city and the date. 3. Get the 31-member GFS ensemble forecast for that city from Open-Meteo's free API. 4. Calculate the probability that the actual high temperature falls into each price bucket. 5. Compare my calculated probability to the current Polymarket price, in cents. 6. If my edge is greater than 8%, place a buy order on the underpriced side using Kelly Criterion sizing, 15% fractional. 7. Cap position size at $50 per trade. 8. Log every trade with timestamp, market ID, forecast, market price, calculated probability, and edge percentage. 9. Auto-resolve closed positions and update P&L. Important: Polymarket markets resolve based on the official airport station for each city, not the city center. Use: NYC = KLGA, LaGuardia Chicago = KMDW, Midway Miami = KMIA LA = KLAX Dallas = KDAL, Love Field Output one clean Python script with a config file for cities, edge thresholds, and position limits.