synthesia
Pass
Audited by Gen Agent Trust Hub on Jul 18, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill mandates a 'human-in-the-loop' process, explicitly requiring human approval for every video before publishing, which prevents autonomous generation of potentially harmful or unauthorized content.
- [SAFE]: The instructions include rigorous safety checks for likeness consent and prohibit the creation of avatars for celebrities, politicians, or competitors without authorization.
- [SAFE]: No suspicious technical patterns were detected, such as shell command execution, dynamic code loading, or attempts to access sensitive system files like credentials or SSH keys.
- [SAFE]: The skill's external dependencies are limited to legitimate enterprise services (Synthesia) and well-known technology providers (OpenAI, Google) for intended media production purposes.
- [SAFE]: The guidance regarding the Sora API sunset is purely informational and does not trigger any conditional or time-delayed malicious logic.
Audit Metadata