breach

SUSPICIOUS: the skill is coherent with its stated red-team purpose and shows no credential theft, installer abuse, or exfiltration path, but it is an offensive-security capability for AI agents and can materially increase attack capacity if misused. Risk is driven by dual-use attack-planning scope, not by hidden behavior or malicious data flows.