image-video-gen
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
- [NO_CODE]: The skill provides descriptive instructions for tool coordination and does not include any executable scripts.
- [INDIRECT_PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it ingests untrusted data from search results to guide generation tools.
- Ingestion points: User-provided descriptions and external data from
byted-web-search(SKILL.md). - Boundary markers: None identified; instructions do not specify delimiters for external content.
- Capability inventory: Access to search and image/video generation tools (SKILL.md).
- Sanitization: No input validation or output sanitization is defined within the workflow instructions.
Audit Metadata