video
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides instructions for reverse-engineering viral video styles by ingesting external content from URLs.
- Ingestion points: External data is processed via tools like
watch-videoandsocial-fetchas described inreferences/edit-anatomy.md. - Boundary markers: The skill incorporates a mandatory "review-once gate" requiring the agent to present a beat sheet for human review before execution.
- Capability inventory: The skill manages video production through tool calls for fetching content and instructions for programmatic rendering using Node.js libraries.
- Sanitization: An "Originality guardrail" explicitly directs the agent to extract only technical editing patterns (grammar) rather than the creative expression or embedded instructions from the source material.
Audit Metadata