handdrawn-live-video-generator

Pass

Audited by Gen Agent Trust Hub on Nov 4, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill contains only instructional text for a creative workflow. It does not include executable code, scripts, network operations, or requests for sensitive file access. All external model references (MiniMax H3) are part of the intended vendor platform functionality.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied scene descriptions and style constraints to generate prompts for a video model, creating a potential injection surface. This is mitigated by behavioral guardrails within the instructions. 1. Ingestion points: User-provided scene ideas and mood descriptions in SKILL.md and SKILL.cn.md. 2. Boundary markers: No specific delimiters are defined to separate user data from generated prompts. 3. Capability inventory: The skill triggers a video generation tool (MiniMax H3) following user confirmation. 4. Sanitization: The instructions include a comprehensive 'prohibited items' list (Step 1) that explicitly forbids the creation of horror, jump scares, and threatening imagery.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 4, 2026, 02:58 AM