stress-test

SUSPICIOUS/HIGH-RISK skill. Its purpose is coherent, but that purpose is to let an AI agent perform offensive security testing and destructive lifecycle actions against production using dynamic shell execution. No clear third-party exfiltration or malware payload is shown, but the operational risk is high and disproportionate for routine agent use.