web-interface-guidelines-review
Pass
Audited by Gen Agent Trust Hub on Aug 24, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill's instructions are purely informative and pedagogical, designed to improve the quality of UI code reviews. No command execution, credential access, or persistence mechanisms are present.
- [EXTERNAL_DOWNLOADS]: The skill references documentation and guidelines from well-known and established technology services, including Vercel's Geist design system (vercel.com), Google's Material Design 3 (m3.material.io), Radix UI primitives (radix-ui.com), and the W3C's WCAG 2.2 recommendations (w3.org). These references are neutral and standard for the skill's stated purpose.
- [PROMPT_INJECTION]: No attempts to override safety filters, extract system prompts, or bypass agent constraints were found. The use of "IMPORTANT" or similar emphasis in the descriptions is consistent with standard instructional language.
- [DATA_EXPOSURE]: No hardcoded credentials, sensitive file paths, or network exfiltration patterns were detected. The skill operates on UI code patterns and design tokens.
- [INDIRECT_PROMPT_INJECTION]: While the skill is designed to process untrusted data (external UI code and design specs), it does not include tools or scripts with capabilities that could be exploited (e.g., no subprocess calls, file writes, or arbitrary network requests). The risk is minimal as the agent is restricted to generating review comments.
Audit Metadata