binary-lifting

SUSPICIOUS: the skill is internally consistent and does not show credential theft or obvious malware, but it enables advanced offensive security workflows and optionally pulls mutable third-party content into the agent context. Main risks are dual-use security capability and external-content integrity, not confirmed malicious behavior.