matlab-uihtml-design

Pass

Audited by Gen Agent Trust Hub on Jul 1, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [SAFE]: The skill generates self-contained HTML/CSS/JS code and explicitly avoids external dependencies or CDN links, reducing the risk of supply chain attacks or data exfiltration.
  • [COMMAND_EXECUTION]: The skill provides instructions for the agent to open a local style gallery using the start (Windows) and open (macOS) commands. This is used for the primary purpose of showing UI styles to the user and is considered safe.
  • [PROMPT_INJECTION]: No prompt injection or safety bypass patterns were detected in the instructions.
  • [DATA_EXFILTRATION]: No network operations, sensitive file access, or hardcoded credentials were found.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 1, 2026, 05:51 AM
Security Audit — agent-trust-hub — matlab-uihtml-design