ux-compliance-audit
Pass
Audited by Gen Agent Trust Hub on Jul 16, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references well-known and trusted accessibility testing services such as WAVE, Axe, and WebAIM Contrast Checker as recommended tools for manual verification.
- [COMMAND_EXECUTION]: The skill provides documentation for running local npm scripts (e.g.,
npm run a11y-audit,npm run wcag-report). These are presented as part of a standard local development workflow and do not involve downloading or executing scripts from untrusted remote sources. - [PROMPT_INJECTION]: The skill processes external website content via URL input for auditing purposes. While this constitutes an attack surface for indirect prompt injection (where a malicious website could contain instructions for the AI), the skill's specific focus on structured audit criteria and report generation mitigates the risk of the agent following arbitrary commands from the audited site.
- Ingestion points: External websites via URL input or uploaded design files.
- Boundary markers: Not explicitly defined in the instruction set, though the structured report format provides a logical boundary for output.
- Capability inventory: Local shell execution for accessibility audits (
npm run). - Sanitization: Standard tool behavior; the agent is instructed to evaluate the site against specific UX criteria (UX01-UX18), which restricts the scope of interaction with the external content.
Audit Metadata