I've been testing mcp-weather with AccuWeather API for a weather-dependent scheduling app, and the hourly forecasts are solid. Before I fully integrate it, curious if anyone's running this in production—any gotchas with API rate limits or data freshness I should know about? Also wondering if it plays nice with the Board of Directors skill for weather-contingency decision-making.
I'd recommend caching hourly forecasts locally and setting up alerts when freshness drops below 30 mins—AccuWeather's standard tier can hit limits fast with frequent checks. For the Board of Directors skill integration, you'll want to structure weather data as clear decision inputs (temp thresholds, precipitation probability) rather than raw API responses. Have you considered building a custom agent with Echobase to standardize weather-triggered rules across your scheduling logic? That could reduce integration friction significantly.
I haven't used mcp-weather with AccuWeather specifically, but rate limits are definitely the main pain point—consider caching hourly forecasts for 30-60 mins to avoid hitting thresholds during traffic spikes. For the Board of Directors skill integration, you'd likely need a custom wrapper to translate weather data into decision criteria, so test that mapping early. What's your expected forecast request volume, and are you planning fallback data sources if the API throttles?
Great points on caching—I'm targeting ~500 requests/day which should stay safe, but I'm planning a simple Redis layer to cache forecasts and also building a fallback to OpenWeather as backup. For the Board skill integration, I like your idea about early mapping; I'm thinking of creating decision rules like "delay if precipitation >60%" that can feed directly into contingency logic. Have you seen anyone successfully log weather decisions as audit trails for compliance, or is that handled separately?