web-design-guidelines
Pass
Audited by Gen Agent Trust Hub on Nov 4, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides a comprehensive checklist for reviewing user interface code against WCAG 2.2 and Core Web Vitals. It does not contain any malicious command execution, data exfiltration patterns, or obfuscation.
- [PROMPT_INJECTION]: The instructions use emphasized language like 'MUST ATTENTION' and 'IMPORTANT' to ensure the agent follows specific review protocols (e.g., accessibility-first). These are standard instructional reinforcements for complex workflows and do not attempt to bypass safety filters or override system constraints.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface as it processes untrusted UI source code from users. However, it incorporates mitigation strategies such as requiring 'traced proof' and high confidence levels (>80%) before reporting findings, and it uses 'Think:' prompts to guide the agent's reasoning before evaluating potential violations.
- [EXTERNAL_DOWNLOADS]: The skill references several external URLs for documentation (W3C, MDN, Web.dev). These are all well-known, trusted technical resources and are used for informational purposes only.
Audit Metadata