ui-ux-pro-max
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: A thorough analysis of the skill's instructions, Python scripts, and data files revealed no evidence of malicious intent, credential theft, or unauthorized network operations.
- [COMMAND_EXECUTION]: The documentation includes prerequisite instructions for the user to install Python using standard system package managers (e.g., brew, apt, winget). These are standard environment setup steps and are not executed automatically by the skill's code.
- [INDIRECT_PROMPT_INJECTION]: The skill exposes a surface for processing untrusted user input as search queries for design patterns. However, the Python implementation safely tokenizes this input via regex and restricts file-writing operations to a specific design-system directory within the agent's workspace.
- Ingestion points: User-provided keywords and page names are passed as arguments to the search.py script.
- Boundary markers: None used in the shell command templates.
- Capability inventory: The skill has the capability to write markdown documentation files to the local file system (design_system.py).
- Sanitization: Input used for search logic is filtered using regex to remove non-alphanumeric characters.
Audit Metadata