conducting-network-penetration-test

SUSPICIOUS/HIGH-RISK skill. The content is internally consistent with its purpose, but that purpose is to equip an AI agent with end-to-end offensive penetration-testing capabilities, including exploitation, credential theft, and lateral movement. Main risk is not deception or exfiltration to a third-party server; it is enabling powerful real-world attack actions against network targets.