knowledge-video-director

Pass

Audited by Gen Agent Trust Hub on Oct 15, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-defined educational goals, evidence, and scripts to generate instructions for downstream video production tools.
  • Ingestion points: User-provided video targets, factual evidence, and knowledge content as described in SKILL.md.
  • Boundary markers: The instructions require explicit user confirmation of the design plan before proceeding to production, although no technical delimiters for data isolation are mentioned.
  • Capability inventory: Generation of storyboard parameters, visual prompts, and execution commands for other local skills like handraw-browser-producer and handraw-jianying-producer.
  • Sanitization: No automated sanitization or escaping of the user-provided data is documented.
  • [SAFE]: The skill operates entirely within a local project context, referencing sibling directories and files using relative paths. It does not perform network operations, use remote code, or request elevated privileges. The use of well-known frameworks like Remotion for video rendering is consistent with the skill's stated purpose.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 15, 2026, 03:42 PM