Ontos-AI

weather

Top source: clawdbot/clawdbot

Ontos-AI 1 Updated 4mo ago
GitHub

Install

npx skillscat add ontos-ai/skills-evaluator/weather

Install via the SkillsCat registry.

SKILL.md

name: weather
description: A skill that can retrieve current weather conditions and forecasts for a specified location.
tags:

  • weather
  • forecast
  • climate
  • meteorology

Weather Skill

This skill provides current weather conditions and forecasts for any given location worldwide. It leverages external weather APIs to fetch up-to-date meteorological data.

Core Functionality

  1. Current Weather: Get real-time weather information including temperature, humidity, wind speed, atmospheric pressure, and general conditions (e.g., sunny, cloudy, rain).
  2. Hourly Forecast: Retrieve weather predictions for the next few hours.
  3. Daily Forecast: Obtain a multi-day weather forecast, typically for 3, 5, or 7 days.
  4. Location Support: Supports various location inputs, including city names, postal codes, and geographical coordinates.
  5. Unit Conversion: Allows users to specify preferred units for temperature (Celsius/Fahrenheit) and wind speed (m/s, km/h, mph).

Usage

Trigger Phrases:

  • "What's the weather like in [Location]?"
  • "Weather forecast for [Location] for tomorrow."
  • "Will it rain in [Location] this weekend?"
  • "Current temperature in [Location]."

Parameters

  • location (Required): The city, postal code, or coordinates for which to fetch weather information (e.g., "London, UK", "90210", "40.7128,-74.0060").
  • type (Optional): Specifies the type of weather information requested. Can be current, hourly, or daily. Defaults to current.
  • days (Optional): For daily forecasts, the number of days to forecast (e.g., 3, 5, 7). Defaults to 3.
  • units (Optional): Preferred units for temperature. Can be celsius or fahrenheit. Defaults to celsius.

Example API Call (Internal)

# Assuming an internal function `get_weather_data`
weather_data = get_weather_data(
    location="Manchester, GB",
    type="daily",
    days=30, # For monthly forecast, assuming API supports up to 30 days
    units="celsius"
)

Output

The skill will return a structured response containing the requested weather information. For forecasts, it will typically include date, high/low temperatures, conditions, and precipitation probability for each period.

{
  "location": "Manchester, GB",
  "type": "daily",
  "forecast": [
    {
      "date": "2023-10-26",
      "min_temp": "8°C",
      "max_temp": "14°C",
      "conditions": "Partly cloudy",
      "precipitation_probability": "20%"
    },
    // ... more days
  ]
}