scan

SUSPICIOUS. The stated purpose matches the offensive scanning behavior, but this is a high-risk penetration-testing skill that enables active scans and SQL injection testing by an AI agent. The main concerns are the offensive capability itself, unreviewed local wrapper scripts, and direct shell interpolation of target input; the pre-execution directive is benign and there is no evidence of credential theft, covert exfiltration, or malicious installer behavior.