kma-weather

What This Skill Does

The kma-weather skill provides OpenClaw agents with direct, real-time access to high-precision weather data from the Korea Meteorological Administration (KMA). It serves as a comprehensive interface for retrieving official current conditions, ultra-short-term forecasts (6 hours), short-term forecasts (up to 3 days), mid-term outlooks (3-10 days), and official government-issued weather warnings. By leveraging KMA's 5km×5km grid system, the agent can provide hyper-local weather updates based on specific latitude and longitude coordinates, ensuring high reliability for users in South Korea.

Installation

To install this skill, use the ClawHub CLI: clawhub install openclaw/skills/skills/steamb23/kma-weather. After installation, you must register on the 공공데이터포털 to obtain an API service key for the KMA's Open API services. Set this key as the KMA_SERVICE_KEY environment variable within your agent's configuration, either in the agents.yaml file for sandbox or host environments, or by exporting the variable directly in your shell session. The skill uses this key to authenticate requests across the short-term, mid-term, and warning API endpoints.

Use Cases

This skill is perfect for agents managing personal schedules, automated smart home controllers, or logistics planning. It allows an agent to automatically check if it is going to rain before a scheduled outdoor delivery, provide daily briefing summaries including expected temperature fluctuations for a user's commute, or trigger alerts if a severe weather warning is issued by the KMA for the user's specific location.

Example Prompts

1. "What is the current weather in Seoul, and is it expected to rain during my commute later today?"
2. "Check the weather forecast for the next 3 days at my coordinates (37.5665, 126.9780) and let me know if I should bring an umbrella."
3. "Are there any active weather warnings for the Gyeonggi province area right now?"

Tips & Limitations

• The KMA API requires exact latitude and longitude for grid-based requests; ensure your agent's location data is accurate.
• The API key is shared across three different KMA service modules, so verify that your key has permissions enabled for all three listed in the setup guide.
• Mid-term forecasts (3-10 days) are based on regional codes rather than exact coordinates, so results for these queries may be less granular than short-term forecast data.
• Frequent polling of the API may hit your daily quota limits; consider setting reasonable cache intervals for your agent's requests.