The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes existing DAG source code and repository files to discover patterns. This creates a potential surface for indirect prompt injection if a processed file contains malicious instructions designed to influence the agent.
Ingestion points: get_dag_source_code and repository file search mentioned in SKILL.md.
Boundary markers: None specified in the workflow.
Capability inventory: The skill has the ability to trigger DAG runs and write/update files.
Sanitization: No explicit sanitization of ingested code is mentioned.
[SAFE]: The skill instructs the agent to use specific Airflow extension tools for environment inspection and operational tasks, which is a standard and recommended practice for this integration.

authoring-dags