penetration-tester

SUSPICIOUS: the skill is internally consistent, but it equips an AI agent for offensive security, phishing, exploitation, and physical/social-engineering operations with only policy-based safeguards. It is not confirmed malware, but it presents high operational and misuse risk.