ui-ux-pro-max
Warn
Audited by Gen Agent Trust Hub on Jul 30, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: SKILL.md instructs the agent to execute shell commands for environment setup, including system-level package installation using various managers (brew, apt, winget).
- [COMMAND_EXECUTION]: SKILL.md provides instructions to use 'sudo' for administrative package management on Linux, which constitutes a privilege escalation vector.
- [COMMAND_EXECUTION]: The Python scripts 'search.py' and 'design_system.py' include functionality to create directories and write markdown files to the local file system when the '--persist' flag is used.
- [COMMAND_EXECUTION]: The persistence logic in 'scripts/design_system.py' relies on user-provided project names to create file paths without sufficient path sanitization, which could potentially allow writing files to unintended locations on the system.
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by ingesting untrusted user input from search queries into persistent 'MASTER.md' files. These files are then prioritized as authoritative design rules in the agent's workflow. This surface is characterized by input ingestion in 'scripts/search.py', the file-writing capabilities in 'scripts/design_system.py', and the absence of boundary markers or input validation.
Audit Metadata