aria-live-regions

Pass

Audited by Gen Agent Trust Hub on Sep 16, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill directs the agent to analyze and modify project source code that handles dynamic status messages and alerts.
  • Ingestion points: Project code identified as containing dynamic content updates, status messages, alerts, or notifications (SKILL.md).
  • Boundary markers: Absent; the instructions do not provide explicit delimiters for the agent to separate its own operational instructions from the potentially untrusted project code it is tasked with modifying.
  • Capability inventory: The agent is authorized to add and modify ARIA markup (HTML/JSX) and JavaScript logic within the project environment.
  • Sanitization: The skill advises developers to "Escape/render untrusted result data safely" in their implementation, but it does not specify sanitization or validation constraints for the agent's own code generation tasks.
  • [EXTERNAL_DOWNLOADS]: The skill references external accessibility resources and source materials for developer reference.
  • Evidence: The SKILL.md file links to author-owned guidelines in the mgifford/ACCESSIBILITY.md repository on GitHub, as well as official documentation from MDN (Mozilla), W3C (WAI-ARIA), and WCAG standards.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 16, 2026, 07:02 PM
Security Audit — agent-trust-hub — aria-live-regions