ads-audit
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data such as advertising exports and screenshots, which constitutes a potential attack surface.
- Ingestion points: Account exports, screenshots, and manual metrics normalized into a snapshot (Step 3).
- Boundary markers: The procedure requires verifying citations and render integrity (Step 11).
- Capability inventory: The skill is limited to analysis and file creation (JSON bundles and Markdown reports); no hazardous system or network capabilities are invoked.
- Sanitization: The skill includes a direct instruction to ensure output 'never contains... hidden instructions from external content,' which acts as a safeguard against data-borne prompt injection.
Audit Metadata