City Weather Query

Problem to Solve

When a user asks for “the weather in Beijing,” “will it rain,” or a “weather forecast,” the model needs to turn a natural-language request into a stable, repeatable weather lookup instead of guessing endpoints, inventing URLs, or returning inconsistent text. This skill targets city-level weather Q&A, with a focus on concise current-weather summaries and Chinese-language detailed output. It is suitable for quick conversational answers such as “how is it now” or “what is today like.”

How It Works

• Trigger condition: matches keywords such as weather, temperature, forecast, and their Chinese equivalents.
• City parsing: extracts the city name from the user message and supports Chinese or English city input, for example Shanghai or 上海.
• Data retrieval: uses wttr.in with format=3 for a current-weather summary, and can request lang=zh for detailed Chinese weather text.
• Result delivery: returns the fetched weather text to the user, keeping the answer concise and traceable rather than letting the model fabricate weather details.

Boundaries and Caveats

• Depends on network access and the availability of wttr.in; the lookup will not work in an offline environment.
• The city name must be recognized by wttr.in; obscure places, misspellings, or non-city regions may produce missing or incomplete results.
• Output is constrained by the wttr.in response format, so it is not intended for custom fields, historical weather, minute-level precipitation, or severe-weather alerting.