web-design-guidelines

Pass

Audited by Gen Agent Trust Hub on Apr 16, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADS
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Fetches fresh guidelines from Vercel Labs' official GitHub repository. This is a legitimate operation and uses a trusted source.
  • [DATA_EXFILTRATION]: No sensitive local files, credentials, or unauthorized network activity were identified.
  • [PROMPT_INJECTION]: No instructions were found that attempt to bypass safety guidelines or override agent behavior.
  • [PROMPT_INJECTION]: The skill ingests external guidelines which could theoretically serve as a vector for instructions. Evidence: 1. Ingestion points: remote content fetched from GitHub (SKILL.md). 2. Boundary markers: absent. 3. Capability inventory: limited to file reading and text-based code analysis. 4. Sanitization: absent.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 16, 2026, 03:10 PM