All products

Query weather data with AI

Turn a place, time and weather question into an Xweather request, check its access cost and choose whether to fetch the results.

To query weather data with AI, decide what would count as an answer. A venue manager asking whether the rain has started needs a recent observation. Someone preparing tomorrow’s event schedule needs a forecast for a particular period. The weather-api skill maps the question to an Xweather endpoint, chooses filters that endpoint supports and prepares a request you can inspect.

A daily summary and a string of hourly values may both be available for a site. Ask for the smaller result first if it answers your question. Before a live call the guide explains the expected access cost. Afterwards it can report the charged amount and the remaining allowance when the service supplies those figures.

What it does

  • Check which station produced an observation and when it was recorded.
  • Specify an hourly or daily forecast for the requested place.
  • Choose an endpoint-specific filter for an alert or an observation.
  • Estimate the access units before fetching a larger response.
  • Inspect an unsuccessful request and distinguish missing credentials from a plan restriction.

Install and use

  1. Install the skill in your agent. The included helper needs Python 3, and live data calls require internet access and entitled Xweather credentials.
  2. Describe the location, relevant time and quantity, then inspect the request with credential placeholders.
  3. Approve the fetch, review the response time and units, and include the provider credit when displaying its data.

FAQ

Source

Made by Vaisala Xweather. Original skill: weather-api, published on GitHub under the MIT licence.

View on GitHub