ETL Pipeline
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion and processing of data from multiple external sources including SQL databases (PostgreSQL, MySQL), NoSQL databases (MongoDB), and third-party APIs (Stripe, Salesforce). This processing creates an attack surface where instructions embedded within the source data could potentially manipulate the agent's logic.
- Ingestion points: Data extraction from external databases and API endpoints as described in the
sourcesandapi_sourcessections ofSKILL.md. - Boundary markers: No explicit delimiters or system instructions to ignore embedded commands within the processed data were defined.
- Capability inventory: The skill possesses powerful capabilities for data manipulation and storage, including the use of tools such as
postgres_query,mysql_query,mongodb_aggregate,bigquery_load, andsnowflake_api. - Sanitization: The instructions do not specify data validation, escaping, or sanitization procedures for external content before it is transformed or loaded into destination warehouses.
Audit Metadata