airflow-dag-patterns
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [SAFE]: The skill provides educational resources and code templates for Apache Airflow development that adhere to industry standards. All code samples consist of standard Python patterns for workflow orchestration.
- [EXTERNAL_DOWNLOADS]: The skill references official documentation and community resources from well-known and trusted sources.
- Evidence: Links to official Apache Airflow documentation and Astronomer guides.
- [INDIRECT_PROMPT_INJECTION]: The templates demonstrate data ingestion from external sources (e.g., reading from S3 via Pandas). While this highlights a potential surface for indirect prompt injection within the resulting pipelines, the templates follow best practices for idempotency.
- Ingestion points:
resources/implementation-playbook.mdcode samples usingpd.read_csvandS3KeySensor. - Boundary markers: N/A (code templates).
- Capability inventory: Python code execution within the Airflow environment and network access via
requests. - Sanitization: None implemented in the illustrative templates.
- [DYNAMIC_EXECUTION]: The skill demonstrates the standard Airflow pattern for dynamic DAG generation using the
globals()dictionary. This is a framework-specific requirement and does not pose a security risk in this context. - Evidence:
resources/implementation-playbook.mdimplementation ofdynamic_dag_factory.py.
Audit Metadata