re-source

SUSPICIOUS. The skill is internally coherent and does not show credential theft or exfiltration, but it grants an AI agent meaningful reverse-engineering capability and persistent binary/project modification. Main concern is offensive-security enablement, not malware behavior.