skills/ses2905/ai-toolkit/taste-skill/Gen Agent Trust Hub

taste-skill

Pass

Audited by Gen Agent Trust Hub on Sep 20, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill acts as a technical reference and style guide for frontend development. It provides specific instructions for aesthetic inference, layout diversification, and performance optimization.
  • [EXTERNAL_DOWNLOADS]: The skill references and provides installation commands for numerous well-known and reputable frontend libraries and design systems, including Motion, Radix UI, Material Web, Fluent UI, and IBM Carbon. These downloads target official package registries (NPM) or established content delivery networks (e.g., Simple Icons).
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied design briefs as its primary input. While this represents a potential injection surface, the skill implements a mandatory "Design Read" confirmation step, requiring the agent to restate its interpretation of the brief before generating code, which effectively serves as a validation boundary.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 20, 2026, 05:58 AM