web-design-guidelines

Pass

Audited by Gen Agent Trust Hub on Apr 14, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Fetches updated interface guidelines from Vercel Labs' official GitHub repository to define the compliance rules.
  • [DATA_EXPOSURE]: Reads local files as specified by the user to perform the web interface review.
  • [PROMPT_INJECTION]: The skill ingests external data (guidelines) which could theoretically contain instructions, however, the source is a trusted organization and the skill acts as a surface for indirect prompt injection rather than containing a direct attack.
  • Ingestion points: SKILL.md (via WebFetch from github.com/vercel-labs)
  • Boundary markers: None
  • Capability inventory: Reads local files
  • Sanitization: None
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 14, 2026, 02:39 AM