ui-ux-pro-max
Fail
Audited by Gen Agent Trust Hub on Jul 15, 2026
Risk Level: HIGHCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The SKILL.md file contains setup instructions advising the agent to use 'sudo' for installing system packages (e.g., 'sudo apt install python3'). This encourages the use of elevated privileges, which is a high-risk practice for AI agents.
- [COMMAND_EXECUTION]: The 'scripts/design_system.py' script contains a path traversal vulnerability in the 'persist_design_system' function. It uses the user-provided 'project_name' and 'page' arguments to construct filesystem paths for directory creation ('mkdir') and file writing ('open') without sufficient sanitization. A malicious input containing traversal characters (e.g., '../../') could allow writing files to arbitrary locations accessible by the agent.
- [PROMPT_INJECTION]: The skill exhibits a significant attack surface for indirect prompt injection.
- Ingestion points: The 'project_name' and 'page' arguments in 'scripts/design_system.py' and the 'query' passed to 'search.py' are used to generate file content and paths.
- Boundary markers: Absent; external content is interpolated directly into generated Markdown files.
- Capability inventory: The skill can create directories and write/overwrite files on the local filesystem.
- Sanitization: Minimal; it only replaces spaces with dashes, which does not prevent directory traversal or malicious content injection.
Recommendations
- AI detected serious security threats
Audit Metadata