setting-up-warehouse-properties
Pass
Audited by Gen Agent Trust Hub on Sep 24, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests external data from data warehouse schemas, which represents a potential attack surface.
- Ingestion points: Column names and types are retrieved from external sources via
external-data-schemas-retrieveandview-get(SKILL.md). - Boundary markers: The instructions mandate that the agent must propose the complete column mapping to the user for review and approval before performing any creation actions (SKILL.md, Step 3).
- Capability inventory: The agent uses
custom-property-definitions-createandcustom-property-sources-createto modify the environment based on the ingested schema data. - Sanitization: The skill provides explicit guidance to exclude sensitive columns (e.g., password hashes, tokens), internal audit columns, and reserved identity properties (e.g., $set, email).
Audit Metadata