image-enhance
Pass
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill incorporates a user-provided 'Subject Name' into output file paths, which creates a surface for directory traversal attacks.
- Ingestion points: User-supplied 'Subject Name' (主题名) in SKILL.md.
- Boundary markers: None provided in the command templates or instructions.
- Capability inventory: File system write operations performed by the local script.
- Sanitization: The skill definition does not describe any validation or sanitization of the user input.
Audit Metadata