airflow-expert
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill documents Airflow DAG patterns that ingest untrusted data from external sources and process it using powerful execution operators, creating a potential surface for indirect prompt injection.
- Ingestion points: Examples in references/CORE_CONCEPTS.md utilize SimpleHttpOperator for API calls, FileSensor for local file system monitoring, and S3KeySensor for cloud storage.
- Boundary markers: No specific boundary markers or delimiters for untrusted data are implemented in the provided code templates.
- Capability inventory: The skill leverages BashOperator for shell execution and PythonOperator for arbitrary code execution in references/CORE_CONCEPTS.md.
- Sanitization: The reference code does not demonstrate sanitization or validation of data fetched from external sensors or operators before it is used in downstream tasks.
- [COMMAND_EXECUTION]: The skill incorporates the use of shell commands through the standard Airflow BashOperator.
- Evidence: references/CORE_CONCEPTS.md contains code for a BashOperator task that executes 'echo "Hello Airflow" && date'.
Audit Metadata