aiwithqasim/weather-airflow-docker-etl — reverse-engineered prompt

Build me a small hands on demo project that shows a complete weather data pipeline running in Docker with Apache Airflow.

It should fetch a free 7 day hourly weather forecast with no API key needed, for Karachi, London, New York, and Tokyo. Then it should save the raw weather data, turn the hourly forecast into daily summary rows, load the final results into a local SQLite database, and print a readable report in the Airflow task logs.

I want it to be classroom friendly, so please include a clear README with setup steps for Docker Desktop, creating the env file, building the containers, starting Airflow, opening the local UI, triggering the DAG, and finding the output files. Use Python for the ETL code and Docker Compose for the Airflow services. Make sure the project creates local output files like raw JSON, a CSV summary, and a SQLite database, and that it works without paid services or API keys.