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) andopen(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