accessibility-review
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data in the form of code files and UI components, which creates a surface for indirect prompt injection if the processed content contains malicious instructions designed to subvert the agent's behavior.
- Ingestion points: The skill identifies and reads code files, components, and design mockups for review, as described in Step 1 of the review process in
SKILL.md. - Boundary markers: Absent; there are no instructions for the agent to use specific delimiters or protective headers when ingesting the target content.
- Capability inventory: The skill has the capability to read local files and modify them ("Implement the fix") during the stepped review workflow defined in
SKILL.md. - Sanitization: No explicit sanitization, filtering, or validation of the ingested code content is performed before processing.
Audit Metadata