frontend-design

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • NO_CODE (SAFE): The skill is purely instructional markdown and does not include any scripts or executable code components.
  • PROMPT_INJECTION (SAFE): Use of keywords like 'CRITICAL' and 'IMPORTANT' are used for styling and design guidance, not for attempting to override agent safety filters or system prompts.
  • DATA_EXFILTRATION (SAFE): No network requests (curl, wget, fetch) or sensitive file path accesses were detected.
  • INDIRECT_PROMPT_INJECTION (LOW): The skill is designed to ingest user-provided frontend requirements to generate code.
  • Ingestion points: User-provided component, page, or application requirements.
  • Boundary markers: None specified in the instructions.
  • Capability inventory: None. The skill generates text/code output but does not possess subprocess, file-system, or network capabilities.
  • Sanitization: None specified.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 04:32 PM