long-description-generation
Pass
Audited by Gen Agent Trust Hub on Sep 24, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill is composed entirely of natural language instructions and documentation intended to guide multimodal models in writing long descriptions for images. No executable scripts or binaries are provided.
- [SAFE]: No network activity, data exfiltration patterns, or hardcoded credentials were identified within the skill content. Image references are relative markdown links to documentation assets.
- [SAFE]: The instructions do not contain prompt injection attempts, persistence mechanisms, or obfuscated content. The instruction to "act as a web accessibility expert" is a benign persona assignment for task performance.
Audit Metadata