accessibility-audit
Pass
Audited by Gen Agent Trust Hub on Nov 10, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill defaults to a read-only mode and explicitly states that the audit does not authorize edits, commits, or publishing without user request.
- [SAFE]: The instructions mandate that source text and page content be treated as untrusted data, providing a defense against potential indirect prompt injection from the content being audited.
- [SAFE]: All external references target official W3C documentation and ARIA pattern guides, which are well-known and trusted authoritative sources.
- [SAFE]: The skill employs restrictive tool usage, limiting evidence collection to context-declared MCP bindings and forbidding the automatic installation of dependencies or remote research.
- [SAFE]: No obfuscation, persistence mechanisms, or unauthorized privilege escalation commands were detected in the skill's instructions or metadata.
Audit Metadata