kawakubo-perspective

Pass

Audited by Gen Agent Trust Hub on Aug 15, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is entirely composed of natural language instructions and structured metadata for character roleplay. There are no scripts, binaries, or executable commands present.
  • [SAFE]: No sensitive file access, credential usage, or network operations were identified. The external GitHub link provided in the footer is for the skill generator tool and is documented neutrally.
  • [SAFE]: The roleplay instructions define a specific persona ("Rei Kawakubo") and include rules for handling user input within that context. These do not attempt to override the underlying AI safety guidelines or extract system prompts.
  • [SAFE]: No obfuscation techniques, hidden text, or malicious prompt injection vectors were found.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 15, 2026, 03:23 AM
Security Audit — agent-trust-hub — kawakubo-perspective