from airflow-vscode-extension43
Manage Apache Airflow from an agent using VS Code extension tools: list, run, debug DAGs, view logs, and check server health — without shell commands.
Provides agent-friendly patterns for operating Apache Airflow through a VS Code extension's toolset. The skill maps common admin intents (list DAGs, trigger runs, inspect logs, pause/unpause) to specific extension tools and offers troubleshooting workflows for failed DAG runs.
Use when an agent is integrated with the Airflow VS Code extension or similar IDE tooling and needs to perform operational tasks (morning health checks, trigger-and-monitor, investigate failures) without invoking CLI/shell commands. Ideal for agents that can call extension-backed tools.
Agents integrated with IDE tooling or extension APIs (VS Code extension backends, Copilot-style integrations, or agents able to call list_active_dags, trigger_dag_run, and analysis tools).
A reference-card skill for the Airflow VS Code extension that maps user intents to 20 extension tools (list/trigger/debug DAGs, view logs, check health). No scripts — purely a routing/lookup table between natural language and tool names. Clean and unambiguous but thin: no error handling guidance, no output contracts, and no guardrails around destructive actions like triggering DAGs or pausing production pipelines.
Pure lookup table skill with no scripts. Security is high as there's no shell execution, no network calls, no credentials. Architecture follows basic skill spec (frontmatter present) but has no scripts/ or references/ dirs and output contracts are undefined. Usefulness is moderate — niche audience (Airflow users with this specific VS Code extension), and the skill is essentially a tool reference card that an agent could infer from the extension's own docs.
