web-design
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill includes instructions to review user-specified files against an accessibility checklist. This creates a surface where an attacker could embed malicious instructions within the files being analyzed to influence the agent's behavior.
- Ingestion points: User-provided design or code files processed during the 'review' workflow in Section 8.
- Boundary markers: The instructions do not specify any delimiters or ignore-instructions warnings to isolate external data from the system prompt.
- Capability inventory: The skill's primary capabilities are limited to generating design tokens (CSS/JS) and text-based guidance. It does not perform subprocess calls, file writes, or network operations.
- Sanitization: No explicit sanitization or validation of the input file content is mentioned.
- [SAFE]: The skill's metadata references a repository from Vercel Labs as an upstream source, which is an established and recognized service provider.
Audit Metadata