analyzing-malware-behavior-with-cuckoo-sandbox

SUSPICIOUS. The skill is internally coherent and does not show credential theft, third-party routing, or deceptive installs, but it gives an AI agent powerful malware-analysis and detonation procedures. Risk is driven by the offensive/security-tool nature of the skill rather than hidden exfiltration or malware behavior in the instructions themselves.