ux-product-auditor
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze external data, such as websites, applications, and onboarding flows. This creates a surface where the processed content (e.g., text or metadata within an interface being audited) could attempt to influence the agent's behavior. However, the skill instructions are strictly diagnostic and do not include capabilities for code execution or network writes based on the processed content.
- Ingestion points: Website URLs, application flows, and design files provided by the user for auditing.
- Boundary markers: Not explicitly defined, but the skill operates within a diagnostic framework.
- Capability inventory: No file system writes, network exfiltration, or dynamic code execution are present.
- Sanitization: Not explicitly implemented for the processed UI content.
- [SAFE]: The skill references several external sources to verify accessibility and privacy compliance. These sources include official US government regulations (Section 508, CCPA), European standards (EN 301 549), and international web standards (W3C WAI-ARIA, WCAG 2.2). All provided links point to established, authoritative domains.
Audit Metadata