race-condition

This skill is coherent with its stated purpose, but that purpose is offensive web exploitation. Its main risk is not hidden malware but enabling an AI agent to conduct high-impact security testing and abuse against targets. Install trust is mixed but not overtly malicious; overall classification is SUSPICIOUS due to exploit capability and autonomous attack guidance.