tracing-upstream-lineage
Pass
Audited by Gen Agent Trust Hub on Feb 25, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [SAFE]: The skill follows best practices for identifying data sources and transformations. It does not contain obfuscation, privilege escalation, or persistence mechanisms.
- [COMMAND_EXECUTION]: The skill instructs the agent to use the 'af' CLI tool (associated with the vendor, astronomer) for diagnostic purposes. Commands such as
af dags list,af dags source, andaf tasks listare used appropriately to gather information about the environment's data pipeline structure. - [PROMPT_INJECTION]: The skill involves the ingestion of external data by reading Airflow DAG source code. While this identifies an indirect prompt injection surface (Category 8), the risk is minimal as the instructions guide the agent to perform structural analysis (e.g., searching for 'FROM' clauses) rather than executing or blindly following instructions contained within the source code.
Audit Metadata