create-dataset-index
Pass
Audited by Gen Agent Trust Hub on Sep 24, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted content from local SQL files and user-pasted query results to generate table profiles.
- Ingestion points: Repository SQL files (Step 2) and user-provided query results in chat (Step 4).
- Boundary markers: The skill does not define specific delimiters or instructions to ignore embedded commands within the ingested data.
- Capability inventory: The skill is restricted to reading local files and writing documentation to the
datasets/directory; it lacks network access and arbitrary code execution capabilities. - Sanitization: There is no evidence of sanitization or validation of the content extracted from SQL files or query results before it is incorporated into the generated markdown documentation.
- [SAFE]: The skill includes an explicit 'hard rule' that prohibits the agent from attempting to connect directly to the user's data warehouse, even if credentials are available. This design choice ensures all database interactions are reviewed and performed by the user, providing a significant security boundary.
Audit Metadata