data-engineering-data-pipeline
Pass
Audited by Gen Agent Trust Hub on Feb 19, 2026
Risk Level: SAFE
Full Analysis
- [SAFE] (SAFE): No malicious patterns, prompt injections, or unauthorized data access mechanisms were identified in the skill instructions or metadata.
- [Indirect Prompt Injection] (LOW): The skill describes an architecture for processing untrusted external data, which is a common surface for indirect prompt injection.
- Ingestion points: External databases (PostgreSQL), Kafka streams, and S3 cloud storage.
- Boundary markers: Absent. The skill does not define specific delimiters for the agent to distinguish between instructions and ingested data.
- Capability inventory: The skill leverages powerful tools including Apache Spark, dbt, and workflow orchestrators (Airflow/Prefect) capable of file and network operations.
- Sanitization: The instructions include best practices such as schema validation and dead letter queues (DLQ) to handle malformed data.
Audit Metadata