Enhanced Weather Trading

Problem

In weather prediction markets, manually tracking NOAA forecasts, finding Simmer markets, and judging liquidity is hard to sustain. This skill turns forecast fetching, city geocoding, market filtering, order execution, and exit rules into a repeatable process, reducing repetitive pricing and calculation work.

How It Works

The skill defaults to autostart:false, so it will not run automatically; when enabled, it executes on a six-hour schedule. It:
- fetches temperature forecasts from api.weather.gov and resolves city names to coordinates via nominatim.openstreetmap.org;
- reads portfolio data, discovers weather markets, and executes trades through api.simmer.markets;
- applies a dynamic confidence range of 60%–90% based on forecast lead time;
- scores market quality using liquidity, volume, time to resolution, and price extremes;
- buys below the entry threshold, sells above the exit threshold, and respects per-trade and per-run limits.

Configuration is typically controlled by SIMMER_API_KEY and SIMMER_WEATHER_* parameters, such as entry price, exit price, maximum position size, target locations, and minimum quality.

Boundaries

This is live automated trading with real funds, not a paper-trading tool. There is no per-trade approval by default, so enabling autonomous mode requires accepting unattended execution risk. Use a least-privilege API key that can read the portfolio and place trades but cannot withdraw funds, then test manually and monitor settlement. It is suitable for engineers who understand prediction markets, the Simmer platform, API risk, and how to review source code and logs.