design-taste-frontend
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill references installation commands for numerous well-known and trusted UI libraries and design systems (e.g., Fluent UI, Material Web, Carbon, Radix). These resources originate from established technology vendors and official public-sector repositories.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided design briefs and external references. It mitigates potential risks through an internal reasoning process: 1. Ingestion points: Section 0.A (Brief inference signals). 2. Boundary markers: The skill requires a mandatory 'Design Read' declaration (Section 0.B) before code generation. 3. Capability inventory: Generates production-ready React/Tailwind code and incorporates visual assets from trusted services. 4. Sanitization: The output is audited against an exhaustive 'Pre-Flight Check' (Section 14) that enforces strict architectural and visual constraints.
- [SAFE]: References to external assets are limited to established iconography CDNs and photography placeholder services. No hardcoded credentials, sensitive file access, or malicious obfuscation patterns were detected.
Audit Metadata