database-security

This is a coherent offensive database-assessment skill rather than a disguised credential harvester, but it gives an AI agent high-risk security testing and potential command-execution workflows against live systems. The main concern is offensive capability and downstream exploitation potential, not malicious data routing or a confirmed supply-chain payload in this fragment.