ai-video-generation
Pass
Audited by Gen Agent Trust Hub on Aug 31, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied data (prompts, image URLs, audio URLs) which are passed as arguments to the
beltCLI for video generation. - Ingestion points: Data enters the context via parameters in the
belt app runcommand withinSKILL.mdexamples. - Boundary markers: Commands use JSON structures to separate prompt text and URLs from execution parameters.
- Capability inventory: The skill uses the
beltcommand-line tool (restricted via frontmatter) to perform network operations and interact with external AI inference services. - Sanitization: No explicit sanitization or input validation logic is defined within the skill instructions.
Audit Metadata