video-gen
Pass
Audited by Gen Agent Trust Hub on Sep 3, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill manages interactions with external video generation APIs (Google Vertex, Seedance, Kling) using a structured approach that prioritizes prompt engineering and visual consistency anchors.
- [SAFE]: External media handling is performed by downloading resources into a sandbox workspace and importing them via platform tools, avoiding direct execution or unsafe handling of remote content.
- [SAFE]: All referenced external documentation URLs point to legitimate, official domains of service providers.
- [SAFE]: Analysis of all threat categories, including prompt injection, obfuscation, data exfiltration, and malicious persistence mechanisms, revealed no security findings.
Audit Metadata