testing-dags

Originally fromastronomer/agents
SKILL.md

DAG Testing Skill

Use the Airflow VS Code extension tools to test, debug, and fix DAGs in iterative cycles.

First Action: Trigger the DAG

When the user asks to test a DAG, your first action should be:

  • trigger_dag_run with the provided DAG ID and config

Do not do pre-flight checks unless the user asks for them.

Testing Workflow Overview

  1. Trigger and monitor
  2. If success, summarize
  3. If failed, debug
  4. Fix and retest

Phase 1: Trigger and Monitor

  1. Trigger: trigger_dag_run
  2. Monitor:
    • Use get_dag_runs to check state
    • If a specific run ID is known, use get_dag_run_detail

Response Interpretation

  • Success: summarize duration and outcome
  • Failed: move to Phase 2
  • Running: ask whether to keep polling

Phase 2: Debug Failures

  1. Use analyse_dag_latest_run for a full analysis of the latest run
  2. If a specific run ID is known, use get_dag_run_detail
  3. If task logs are needed, open go_to_dag_log_view

Phase 3: Fix and Retest

  1. Apply fixes in the DAG or related systems
  2. Re-trigger with trigger_dag_run
  3. Re-check with get_dag_runs or get_dag_run_detail

Notes

  • Prefer analyse_dag_latest_run for fast triage
  • Use get_dag_history when comparing across days
Weekly Installs
18
GitHub Stars
42
First Seen
Feb 6, 2026
Installed on
github-copilot17
opencode4
cursor4
gemini-cli3
codebuddy3
codex3