create-podcast-skit-ad
Pass
Audited by Gen Agent Trust Hub on Jul 10, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill incorporates explicit 'human gates' at critical stages of the workflow (idea-brief lock, script approval, and final quality check). This design pattern significantly reduces the risk of malicious or unintended content being processed without oversight.
- [SAFE]: Interaction with external AI providers such as ElevenLabs, Veed, and fal.ai is consistent with the skill's primary purpose. No sensitive data exfiltration or credential harvesting patterns were observed.
- [SAFE]: The inclusion of specific voice IDs and sample brand URLs are for demonstration and template purposes and do not represent a security risk.
- [SAFE]: While the skill provides tips on adjusting prompts to avoid NSFW filter rejections (e.g., changing 'narrowed eyes' to 'curious questioning look'), these are presented as creative troubleshooting rather than an attempt to bypass agent safety protocols or jailbreak the system.
Audit Metadata