web-cloud-designer

Pass

Audited by Gen Agent Trust Hub on Mar 6, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: Analysis of the skill instructions and metadata confirms the absence of any override markers or safety bypass attempts.
  • [DATA_EXFILTRATION]: No patterns for reading sensitive files or exfiltrating data were identified. The included GPU detection logic is a standard implementation for device-based performance scaling.
  • [REMOTE_CODE_EXECUTION]: The skill does not facilitate the downloading or execution of untrusted remote scripts.
  • [COMMAND_EXECUTION]: While the Bash tool is listed as allowed, the skill files contain no shell commands or executable scripts that pose a risk to the host environment.
  • [SAFE]: The skill is focused purely on UI/UX design components and uses well-documented web standards (SVG, CSS3, JavaScript) to achieve its goals.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 6, 2026, 12:53 PM