ai-video-gen
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user data in the form of text prompts and reference image URLs which are passed to external video generation APIs. This creates a vulnerability surface for indirect prompt injection where malicious instructions embedded in the input could attempt to subvert the behavior of the downstream AI models.
- Ingestion points:
input.prompt,input.reference_image_url, andinput.tail_image_urlwithinSKILL.md. - Boundary markers: No explicit delimiters or instructions to ignore embedded commands are included in the prompt handling logic.
- Capability inventory: The skill leverages network capabilities via
curl,fetch, and the Pythonrequestslibrary to interact with external API endpoints such asapi.heygen.com. - Sanitization: No input validation or sanitization mechanisms are demonstrated in the provided code examples.
Audit Metadata