web-design-guidelines

Pass

Audited by Gen Agent Trust Hub on Mar 27, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADS
Full Analysis
  • External Resource Ingestion: The skill retrieves guidelines from a public GitHub repository. This is used to define the rules for the design audit, ensuring the agent uses the latest best practices.
  • Indirect Prompt Injection Surface: The skill exhibits an indirect prompt injection surface because it interprets instructions provided in the fetched external file.
  • Ingestion points: The URL https://raw.githubusercontent.com/vercel-labs/web-interface-guidelines/main/command.md is accessed via the WebFetch tool.
  • Boundary markers: There are no explicit delimiters or warnings defined to separate fetched instructions from the agent's core logic.
  • Capability inventory: The skill is authorized to read local files provided by the user and output formatted text findings.
  • Sanitization: No explicit sanitization or validation of the fetched content is described before it is processed by the agent. While this creates a theoretical opportunity for the external content to influence agent behavior, the use of a well-known technology provider's repository as the source makes this a standard functional pattern.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 27, 2026, 10:08 AM