reverse-eng-malware

SUSPICIOUS: the skill is internally coherent for malware analysis and does not show clear credential theft or third-party exfiltration, but it grants an AI agent advanced offensive/security-research capabilities including unpacking, anti-analysis bypass, and live detonation workflows. The only flagged network flow is a documented localhost CAPE upload, so malware intent is not confirmed; overall risk is high because this skill meaningfully expands agent capability into malware reverse engineering.