brand-video-generator
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructions direct the agent to 'Inspect supplied materials and relevant brand pages' and 'Collect the evidence that affects the concept'. This ingestion of untrusted external data (websites and files) creates a vulnerability surface where malicious instructions could be embedded in the source material.
- Ingestion points: Processes user-supplied assets and external website content as defined in the 'Build an evidence base' section of SKILL.md.
- Boundary markers: No specific delimiters or instructions to ignore embedded commands are present in the skill body.
- Capability inventory: The skill leverages video composition and motion-design skills for implementation, but does not show evidence of dangerous tools like arbitrary command execution or network exfiltration.
- Sanitization: The skill does not define methods for sanitizing or escaping the data retrieved from external brand pages.
Audit Metadata