frontend-accessibility-best-practices
Pass
Audited by Gen Agent Trust Hub on Feb 15, 2026
Risk Level: LOW
Full Analysis
- [SAFE] (SAFE): Comprehensive review of all 8 files indicates the skill is purely instructional. It contains markdown documentation and React/TypeScript code snippets for implementing semantic HTML, ARIA live regions, focus management, and keyboard navigation.
- [Category 2: Data Exposure] (SAFE): No sensitive file paths, environment variables, or hardcoded credentials were found. The code examples use localized strings (e.g.,
t("Close")) and standard UI properties. - [Category 4: Unverifiable Dependencies] (LOW): The skill references standard, reputable libraries such as
react-ariaand@mantine/hooksin its code examples. These are trusted sources for UI accessibility and do not pose a threat in this context. - [Category 8: Indirect Prompt Injection] (LOW): The skill is designed to guide an agent during code reviews or component creation. While this involves processing untrusted user code (the ingestion point), the skill itself does not implement any logic that could be hijacked to perform malicious actions. It lacks network access, file system modification capabilities, or command execution logic.
Audit Metadata