Air Quality Skill

Air quality is a daily decision input — run outside or inside, windows open or closed, mask or not — and Open-Meteo serves it globally over plain HTTPS, no key, no signup. This skill fetches it, then does the part the raw number doesn't: translating PM2.5 and AQI into the standard health bands, matched to what the user was actually deciding.

What This Skill Produces

• The read — one sentence: the air right now and what it suggests for the stated activity
• The numbers — AQI plus the pollutants that matter (PM2.5, PM10, ozone, NO₂), with the timestamp
• The bands — where today sits on the standard scale, so the number has meaning
• The command — the exact curl, rerunnable

Required Inputs

Ask for these if not provided:

• Location — lat/lon or a place name (geocode first: https://geocoding-api.open-meteo.com/v1/search?name=Delhi&count=1)
• The decision behind the question — a run, a bike commute, a sensitive-lungs household, open windows — the read is calibrated to it
• Which index they think in — US AQI or European AQI (the API serves both; the numbers differ substantially for the same air)

Framework: The Fetch and the Read