ui-ux-pro-max
Pass
Audited by Gen Agent Trust Hub on Apr 5, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local Python script (
search.py) to generate design systems and guidelines. It explicitly instructs the agent to sanitize user-provided keywords by stripping or rejecting shell metacharacters (e.g.,;,&&,|,$()) to prevent command injection. - [EXTERNAL_DOWNLOADS]: Provides instructions to install Python 3 using official system package managers, including Homebrew for macOS, APT for Debian/Ubuntu, and Winget for Windows, ensuring dependencies are sourced from established platform providers.
- [COMMAND_EXECUTION]: Utilizes system commands such as
python3 --versionandsudo apt updatefor environment verification and package management. - [PROMPT_INJECTION]: Establishes a security boundary by mandating that the agent treat all user-supplied search terms and CLI tool outputs strictly as data, preventing the interpretation of external content as instructions or behavioral overrides.
Audit Metadata