ui-pattern-research
Pass
Audited by Gen Agent Trust Hub on Oct 24, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill references established, high-reputation external documentation sources (Cloudscape Design, W3C, and Carbon Design System) for UI patterns, ensuring the agent relies on trusted industry standards.
- [SAFE]: Data integrity is prioritized by requiring the agent to verify SHA256 hashes for project profiles, design briefs, and context files before applying recommendations.
- [SAFE]: The instructions establish a clear privacy boundary by explicitly stating that internal project materials must not be uploaded during external search operations.
- [SAFE]: The skill provides defensive instructions against unauthorized environment modifications, such as advising the agent not to automatically install tools if the required CLI is missing.
- [SAFE]: It utilizes a vendor-specific tool ('phoenix-ui') for context gathering in a controlled manner, emphasizing the verification of 'provenance' and 'accepted decisions' rather than blind execution of output.
Audit Metadata