airflow-dag-patterns
Pass
Audited by Gen Agent Trust Hub on Sep 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill describes patterns for ingesting and processing data from external sources (such as S3 buckets using
pandas.read_csv), which constitutes a potential attack surface for indirect prompt injection if the ingested data contains malicious instructions targeting the AI agent. - Ingestion points:
extracttasks inSKILL.mdare designed to retrieve data from external storage (S3) and file systems. - Boundary markers: The provided code templates do not include specific delimiters or instructions to ignore commands embedded within the processed data.
- Capability inventory: The skill uses
PythonOperatorfor arbitrary code execution andrequestsfor network communication within the Airflow environment. - Sanitization: As a template for educational purposes, the examples do not include data validation or sanitization logic.
Audit Metadata