web-design-guidelines
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- EXTERNAL_DOWNLOADS (LOW): The skill fetches markdown content from
https://raw.githubusercontent.com/vercel-labs/web-interface-guidelines/main/command.md. Per the [TRUST-SCOPE-RULE], this is downgraded to LOW becausevercel-labsis a trusted GitHub organization. - PROMPT_INJECTION (LOW): The skill demonstrates an indirect prompt injection surface (Category 8) by fetching a remote file that defines rules and output formats for the agent to follow.
- Ingestion points: Remote instructions fetched from the Vercel GitHub repository via WebFetch.
- Boundary markers: Absent; the skill does not use delimiters to isolate the fetched instructions from the system prompt.
- Capability inventory: The skill reads local user files and performs network fetches.
- Sanitization: Absent; the content from the remote URL is applied directly to the agent's reasoning process.
Audit Metadata