animate-design
Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external Figma links and user-uploaded assets such as photos, logos, and video clips (SKILL.md, references/asset-routes.md). No explicit boundary markers or delimiters are defined to isolate these inputs from the model's instructions. The skill's capability inventory includes modifying scene documents via picsart_scene_patch and executing credit-gated generative tools via picsart_generate (SKILL.md, references/video-generation.md). While the skill requires human-in-the-loop approvals for all changes, no explicit sanitization of ingested data is described.
Audit Metadata