ui-validator

Pass

Audited by Gen Agent Trust Hub on Apr 1, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious behavior or high-risk patterns detected.
  • [COMMAND_EXECUTION]: The skill executes local commands like pnpm dev, lsof, and Test-NetConnection to manage the local development server on port 3000. These are standard developer operations and appropriately scoped.
  • [DATA_EXFILTRATION]: Browser automation actions are confined to localhost:3000. No sensitive data access or external network exfiltration was found.
  • [PROMPT_INJECTION]: No direct prompt injection found. Regarding potential indirect injection: 1. Ingestion points: DOM content and styles from localhost:3000 in SKILL.md. 2. Boundary markers: Absent. 3. Capability inventory: Command execution and browser control tools. 4. Sanitization: Absent. This surface is considered low risk as it is restricted to the local development environment.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 1, 2026, 11:54 AM
Security Audit — agent-trust-hub — ui-validator