ux-best-practices
Pass
Audited by Gen Agent Trust Hub on Nov 13, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill fetches interface guidelines from the official Vercel Labs GitHub repository (vercel-labs/web-interface-guidelines). This is used as a reference for code-level UX reviews.
- [INDIRECT_PROMPT_INJECTION]: The skill analyzes untrusted data from user files and external URLs, creating a surface for indirect prompt injection. Ingestion points: The skill audits user-provided files and fetches markdown data from GitHub. Boundary markers: Explicitly instructs the agent to treat web pages as data and never follow instructions found in them. Capability inventory: The skill is capable of reading local files and making network requests to obtain audit criteria. Sanitization: Uses instructional boundaries to prevent the agent from obeying instructions embedded in the audited data.
Audit Metadata