designer-uiux-specialist

Pass

Audited by Gen Agent Trust Hub on Jul 28, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it is designed to ingest and analyze untrusted external data from rendered web pages.
  • Ingestion points: Rendered frontend pages and code mockups loaded via browser automation tools (SKILL.md).
  • Boundary markers: The instructions do not provide delimiters or specific directives to the agent to ignore instructions that might be embedded in the HTML or content of the target pages.
  • Capability inventory: The skill utilizes computer-use automation and the evaluate_script function within Chrome DevTools MCP.
  • Sanitization: No sanitization or escaping mechanisms are described for the content extracted from external pages before it is processed by the agent.
  • [COMMAND_EXECUTION]: The workflow involves the dynamic execution of JavaScript on external pages using the evaluate_script tool.
  • Evidence: The skill instructs the agent to use window.getComputedStyle(element) to audit CSS properties on live implementations (SKILL.md). While intended for design verification, this allows the agent to execute code within the context of potentially untrusted third-party web content.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 28, 2026, 01:03 PM
Security Audit — agent-trust-hub — designer-uiux-specialist