authority-ad
Pass
Audited by Gen Agent Trust Hub on Sep 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from user briefs and external media analysis to generate instructions for production tools.
- Ingestion points: The skill reads user-supplied product details, brand briefs, and observable evidence from external ads via the
$media-analysistool (SKILL.md). - Boundary markers: It employs 'Non-Imitation Boundary' and 'Authority Integrity' instructions to limit the influence of external content on output behavior (SKILL.md, references/proof-and-claims.md).
- Capability inventory: Upon user approval, the skill hands off scripts to
$image-batch-runnerand$video-batch-runnerfor asset generation (SKILL.md, references/workflow-handoff.md). - Sanitization: It mandates translating specific references into generic production attributes and verifying claims against approved context before processing (SKILL.md).
Audit Metadata