video
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill generates video prompts based on user-supplied creative concepts and scripts, representing a potential ingestion surface for untrusted content.
- Ingestion points: User ideas, scripts, and fixing existing prompts as triggered by keywords in SKILL.md.
- Boundary markers: Relies on the default system-level separation of developer vs user instructions.
- Capability inventory: Text generation only; the skill provides no tools for file manipulation or network access.
- Sanitization: The skill's 'Details Law' and shot cards require the conversion of input into technical cinematographic parameters, which helps structure and sanitize the final output.
- [NO_CODE]: The skill consists exclusively of markdown documentation and reference files. It does not attempt to install or execute any code or external scripts.
- [SAFE]: All external URLs relate to the author's professional resources such as GitHub and personal websites. These are provided for reference purposes and do not trigger automated network activity.
Audit Metadata