ui-ux-pro-max
Warn
Audited by Gen Agent Trust Hub on Mar 30, 2026
Risk Level: MEDIUMPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The persistence module in scripts/design_system.py contains a path traversal vulnerability in the persist_design_system function. The project_name variable is used to construct directory and file paths using the pathlib library's join functionality without sanitizing for directory traversal characters (../) or absolute paths. This allows an agent, if manipulated by a malicious prompt, to write files to arbitrary locations on the filesystem.
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by ingesting and acting upon instructions from data/ui-reasoning.csv.
- Ingestion points: data/ui-reasoning.csv is loaded by scripts/design_system.py.
- Boundary markers: Absent; data values are directly used to influence reasoning logic.
- Capability inventory: File system write operations via the project persistence flag.
- Sanitization: None identified for processed data entries.
Audit Metadata