conducting-social-engineering-penetration-test

SUSPICIOUS/HIGH-RISK skill. Its stated purpose is offensive social-engineering testing, and the capabilities are aligned with that purpose, but the footprint is inherently dangerous for an AI agent: credential harvesting, MFA/session interception, phishing/vishing execution, and physical intrusion tactics. The main issue is not inconsistency but that this skill equips the agent for real-world offensive security operations with clear data-capture paths to operator infrastructure.