This covers the practical side of getting Airflow code into production, whether you're on Astronomer's managed Astro platform or running open source on Docker Compose and Kubernetes. The Astro section is especially useful because it breaks down when to use dag-only deploys versus full image builds, which matters when you're trying to iterate quickly without waiting for Docker builds. The Docker Compose setup walks through the Airflow 3 architecture changes, including the new API server and standalone DAG processor. If you're setting up CI/CD or just trying to understand the difference between pushing DAGs and deploying a full project, this gives you the commands and context to make that decision.
npx skills add https://github.com/astronomer/agents --skill deploying-airflow