frontend-design
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOW
Full Analysis
- [Prompt Injection] (SAFE): No attempts to bypass safety filters or ignore system instructions. Instructional language is limited to design guidance.
- [Data Exposure] (SAFE): No sensitive file paths, credentials, or network exfiltration patterns identified.
- [External Downloads] (SAFE): No remote code execution or package installation commands. References to libraries like Motion are for output context.
- [Indirect Prompt Injection] (INFO): The skill interprets user requirements to generate code. While it lacks sanitization or boundary markers, it has no side-effect capabilities (e.g., file-write, network access). Ingestion point: User frontend requirements. Capability inventory: UI code generation.
Audit Metadata