web-design-guidelines

Warn

Audited by Gen Agent Trust Hub on Feb 28, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill contains deceptive metadata regarding its origin.
  • Evidence: The YAML metadata identifies the author as "vercel", which is inconsistent with the actual author "darraghh1". This discrepancy can mislead users into trusting the skill based on a false affiliation.
  • [EXTERNAL_DOWNLOADS]: The skill fetches design guidelines and instructions from the "vercel-labs" organization on GitHub via a direct URL.
  • Note: This reference targets a well-known service and is documented here neutrally as part of the skill's core functionality.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to the lack of isolation between data sources and instructions.
  • Ingestion points: The skill retrieves rules from a remote GitHub file (command.md) and reads user-provided UI code files.
  • Boundary markers: There are no delimiters or instructions provided to ensure the agent disregards potentially malicious instructions embedded in the external content or the files being audited.
  • Capability inventory: The skill has the ability to read local files and perform network requests using WebFetch.
  • Sanitization: No sanitization, escaping, or validation steps are performed on the content fetched from the network or the local files before they are processed by the agent.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 28, 2026, 02:29 PM