weather

name	weather
description	Query current weather and forecasts for any place using Open-Meteo through curl + jq via the bash tool — no API key and no extra Python dependencies. Two APIs, both key-free: the Geocoding API (geocoding-api.open-meteo.com/v1/search) turns a place name like 'Tokyo' or 'Shibuya' into latitude/longitude/timezone, and the Forecast API (api.open-meteo.com/v1/forecast) returns current conditions, an hourly forecast, and a daily forecast for those coordinates. Weather condition is a WMO weather_code integer that this skill maps to a human label. Use when the user asks what the weather is like now, the temperature/wind/humidity/precipitation at a location, or a forecast (today, hourly, or multi-day) for a city or coordinate. Needs curl + jq installed. Present only values copied verbatim from a real API response — never invent weather data. For plain geocoding/routing without weather use the maps skill instead.
category	research

Weather (当前天气 / 预报查询)

Query current weather and forecasts for any place on earth via Open-Meteo — a free, key-free weather API. Entirely curl + jq through the bash tool. No dedicated Python tool and no extra Python dependencies — it is shell only.

Reply in Chinese unless the user clearly uses another language.

Two APIs, both free and needing no API key:

Need	Method + endpoint
Place name → coordinates + timezone	GET https://geocoding-api.open-meteo.com/v1/search
Current weather / hourly / daily forecast	GET https://api.open-meteo.com/v1/forecast

The typical flow is two calls: geocode the place name to get latitude/longitude/timezone, then pass those to the forecast endpoint. If the user already gives you a coordinate, skip straight to the forecast call.

Related skill: for geocoding, POIs, routing, and timezone lookups without weather, use the [[maps]] skill (OpenStreetMap, also key-free). Use weather when the user wants actual weather or forecast data.

Prerequisites

• curl and jq installed. Check with command -v curl jq. On Windows the bundled msys64 provides both; if jq is missing, install it (winget install jqlang.jq / choco install jq / apt-get install jq).
• No API key and no account needed for non-commercial use. Open-Meteo is free for reasonable request volumes — don't hammer it in tight loops.

command -v curl jq >/dev/null || { echo "need curl + jq"; exit 1; }

Never fabricate weather data (hard rule)

Every temperature, wind speed, humidity, precipitation figure, condition, and forecast time you give the user must come from an actual API response you just received — copy it verbatim from the curl | jq output. Do not guess or invent weather from memory or from what is "typical" for a place or season.

If you can't run a real query — curl/jq missing, network blocked, an HTTP error, an "error": true body, or an empty results array from geocoding — then say so plainly and stop. Report the exact blocker (which endpoint, the HTTP status / reason message, or that geocoding returned no match) and what the user should try. Never substitute made-up weather to fill the gap.

Step 1 — Geocode: place name → coordinates (/v1/search)

Turn a free-text place name (or postal code) into latitude, longitude, and the location's timezone. Only name is required. Useful optional params: count (results, default 10), language (e.g. zh), countryCode (ISO-3166 alpha-2 filter).

PLACE="Tokyo"
curl -s -G "https://geocoding-api.open-meteo.com/v1/search" \
  --data-urlencode "name=$PLACE" \
  --data-urlencode "count=5" \
  --data-urlencode "language=zh" \
| jq -r 'if .results then .results[]
    | "\(.name)\t\(.admin1 // "-")\t\(.country // "-")\t\(.latitude),\(.longitude)\t\(.timezone)"
  else "ERROR / no match: \(.reason // "empty results")" end'

Each result gives .name, .latitude, .longitude, .timezone (a tz-database name like Asia/Tokyo), plus .country, .admin1 (region/state), .population, .feature_code. Pick the match the user meant (disambiguate by country/region if several) and carry its latitude, longitude, and timezone into step 2. Note: a 2-character name only matches exactly; 3+ chars do fuzzy matching. For the Americas, longitudes are negative (west of Greenwich).

Step 2 — Forecast: coordinates → weather (/v1/forecast)

One endpoint serves current conditions, hourly, and daily forecasts — you choose which by sending the current, hourly, and/or daily parameters (each a comma-separated list of variables, no spaces). Required: latitude, longitude. timezone is required whenever you request daily — pass the tz name from geocoding, or auto to resolve it from the coordinates. Other useful params: forecast_days (0–16, default 7), past_days (0–92), temperature_unit (celsius/fahrenheit), wind_speed_unit (kmh/ms/mph/kn).

Current weather

LAT=35.6895; LON=139.6917; TZ="Asia/Tokyo"
curl -s -G "https://api.open-meteo.com/v1/forecast" \
  --data-urlencode "latitude=$LAT" \
  --data-urlencode "longitude=$LON" \
  --data-urlencode "current=temperature_2m,relative_humidity_2m,apparent_temperature,weather_code,wind_speed_10m,precipitation" \
  --data-urlencode "timezone=$TZ" \
| jq -r 'if .error then "ERROR: \(.reason)" else .current as $c |
    "时间: \($c.time)",
    "气温: \($c.temperature_2m)\(.current_units.temperature_2m)  体感: \($c.apparent_temperature)\(.current_units.apparent_temperature)",
    "天气: \($c.weather_code) (见 WMO 对照表)",
    "湿度: \($c.relative_humidity_2m)%",
    "风速: \($c.wind_speed_10m)\(.current_units.wind_speed_10m)",
    "降水: \($c.precipitation)\(.current_units.precipitation)"
  end'

Daily forecast (multi-day)

LAT=35.6895; LON=139.6917; TZ="Asia/Tokyo"
curl -s -G "https://api.open-meteo.com/v1/forecast" \
  --data-urlencode "latitude=$LAT" \
  --data-urlencode "longitude=$LON" \
  --data-urlencode "daily=weather_code,temperature_2m_max,temperature_2m_min,precipitation_sum,precipitation_probability_max,wind_speed_10m_max" \
  --data-urlencode "timezone=$TZ" \
  --data-urlencode "forecast_days=7" \
| jq -r 'if .error then "ERROR: \(.reason)" else .daily as $d |
    range(0; ($d.time|length)) as $i |
    "\($d.time[$i])\t\($d.temperature_2m_min[$i])~\($d.temperature_2m_max[$i])°C\t降水 \($d.precipitation_sum[$i])mm (\($d.precipitation_probability_max[$i])%)\tWMO \($d.weather_code[$i])\t风 \($d.wind_speed_10m_max[$i])"
  end'

Hourly forecast

Request hourly=temperature_2m,precipitation,precipitation_probability,weather_code,wind_speed_10m. The response arrays in .hourly are index-aligned with .hourly.time; slice to the range the user cares about (e.g. next 12 hours) rather than dumping all 168. Use forecast_days=1 to keep it to today.

Common variables — current/hourly: temperature_2m, apparent_temperature, relative_humidity_2m, precipitation, precipitation_probability, weather_code, cloud_cover, wind_speed_10m, wind_direction_10m, is_day. daily: weather_code, temperature_2m_max/_min, apparent_temperature_max/_min, precipitation_sum, precipitation_probability_max, wind_speed_10m_max, sunrise, sunset, uv_index_max.

WMO weather_code → label

weather_code is a WMO 4677 integer. Map it for the user:

Code	天气	Code	天气
0	晴 Clear	51/53/55	毛毛雨 弱/中/强 Drizzle
1	大部晴 Mainly clear	56/57	冻毛毛雨 Freezing drizzle
2	多云 Partly cloudy	61/63/65	雨 小/中/大 Rain
3	阴 Overcast	66/67	冻雨 Freezing rain
45/48	雾 / 雾凇 Fog	71/73/75	雪 小/中/大 Snow
77	米雪 Snow grains	80/81/82	阵雨 弱/中/强 Rain showers
85/86	阵雪 Snow showers	95	雷暴 Thunderstorm
96/99	雷暴伴冰雹 Thunderstorm w/ hail

Present results

Give the user the concrete values copied verbatim from the parsed output: the place (name

• region/country you geocoded), the local time of the reading, temperature (and apparent temperature), condition (WMO code translated via the table above), humidity, wind, and precipitation — with the units from *_units in the response. For forecasts, present a compact per-day or per-hour table. State the timezone the times are in. Credit the source: data from Open-Meteo (open-meteo.com), CC BY 4.0. If geocoding returns no match or the forecast call errors / is blocked, say so and suggest a fix (spell the place differently, add a country, or pass a coordinate directly) — never invent weather to fill the gap.