ui-ux-pro-max
Fail
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: HIGHCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- Privilege Escalation (HIGH): The
SKILL.mdfile contains explicit instructions for the agent to executesudo apt installto install dependencies. Commands utilizing superuser privileges are high-severity risks as they can be manipulated to compromise the host system. - Indirect Prompt Injection (LOW): The skill reads data from local CSV files and presents it to the agent, creating a surface for embedded instructions to influence agent behavior.
- Ingestion points: Multiple CSV files in the
data/directory loaded byscripts/core.py. - Boundary markers: Absent. Search results are interpolated directly into the response without delimiters.
- Capability inventory: Execution of local python scripts and high-privilege installation commands (
apt,brew,winget). - Sanitization: None. Data from CSV files is processed and printed without validation.
- External Downloads (LOW): The skill suggests using system package managers (
apt,brew,winget) to download and install Python. While these are standard tools, the instruction involves external software retrieval.
Recommendations
- AI detected serious security threats
Audit Metadata