linux-lateral-movement

High-risk offensive security skill. Its purpose is honest, but it explicitly instructs an AI agent to steal credentials, pivot to other hosts, escalate privileges, and install persistence, making it unsuitable for general agent use and effectively malicious in operational impact.