This repo contains experimental utilities to help users of Cloud Composer.
For samples, please see Python Docs Samples which contains code samples found in Composer documentation
These tools are NOT under any kind of SLO or SLA and have limited support.
Takes in two Managed Airflow environments and compares their attributes
Takes in a Managed Airflow environment and analyzes the DAGs for compatibility with Composer 2
Profiles DAG parsing for a Managed Airflow environment. This tool helps you optimize parsing latency, including top-level code detection
An administration dashboard to manage DAGs (pause, unpause, trigger, and bulk operations), view import errors, and edit DAG code in-browser across multiple Composer environments, projects, and regions from a single workspace.
Provides production-grade Airflow Cluster Policies for Cloud Composer to enforce resource governance, clamp excessive KubernetesPodOperator requests, and enforce task and DAG metadata standards.
A template and sample configuration for establishing a CI/CD pipeline for Cloud Composer. It features:
- Linting & Formatting: Checks using Ruff.
- Automated Testing: Runs unit and integration tests against local Airflow instances inside a Composer-matching environment.
- Deployment: Automatically syncs DAGs, data files, and dependencies (
requirements.txt) to Cloud Composer environments upon successful validation. - Agentic Remediation: An optional setup using Antigravity CLI to automatically analyze, fix, and propose PRs for DAG failures or optimizations.