product-design-critic

Pass

Audited by Gen Agent Trust Hub on Apr 12, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: References official documentation from trusted organizations including NIST, W3C, Apple, and Microsoft to anchor product critiques.
  • [COMMAND_EXECUTION]: Instructs the agent to run a local script (python3 scripts/record_skill_feedback.py) to persist user feedback for system improvement.
  • [PROMPT_INJECTION]: Contains simulated attack strings in the references/evals.yaml file; these are used for safety evaluation and are not part of the operational instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 12, 2026, 01:14 PM
Security Audit — agent-trust-hub — product-design-critic