What problem does it solve? An AI model's training data contains no knowledge of today's weather, so answering forecast questions from memory produces invented data. This Skill forces every weather answer through live network sources so users get real, current conditions with a named source. ## Core Features & Use Cases - Fast one-line forecasts: Query wttr.in with format parameters for instant current conditions, temperature, and wind for any city. - Full multi-day forecasts: Retrieve three-day forecasts in three-hour slots when users ask about tomorrow or the weekend. - Search fallback path: When wttr.in fails or official warnings are needed, search meteorological services like meteofrance.com or weather.gov and fetch the page. - Use Case: A user asks "what is the weather in Paris this weekend?" The Skill fetches the full wttr.in forecast, extracts the numbers, and reports highs, lows, and precipitation in French with the source named. ## Quick Start Ask the assistant what the weather will be in a specific city and it will fetch the live forecast from wttr.in and report conditions with the source.