web-design-guidelines
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS] (LOW): The skill fetches fresh guidelines from 'https://raw.githubusercontent.com/vercel-labs/web-interface-guidelines/main/command.md'. Per the [TRUST-SCOPE-RULE], this source (vercel-labs) is a trusted organization, which downgrades the risk level for the external reference.
- [INDIRECT_PROMPT_INJECTION] (LOW): The skill creates a surface for indirect prompt injection as it processes both remote content and local user files. 1. Ingestion points: Fetched 'command.md' and local UI files (React, Vue, HTML, etc.). 2. Boundary markers: No explicit delimiters are specified for the external guidelines. 3. Capability inventory: Read local files and generate text analysis; no subprocess or network write capabilities are invoked during the audit. 4. Sanitization: None present. The risk remains low due to the trusted nature of the remote guideline source.
- [DATA_EXFILTRATION] (SAFE): The skill reads local source code files for analysis. However, there are no network post operations or exfiltration patterns detected that would suggest data is being sent to an unauthorized third party.
Audit Metadata