security-arsenal

This skill is purpose-aligned as a red-team arsenal, but that purpose is inherently high risk for an AI agent: it equips the agent to probe, exploit, brute-force, bypass authentication, and exfiltrate data using attacker-controlled and known interception endpoints. It is not confirmed malware because it does not target the user's own machine or hide behavior, but it is a high-risk offensive capability set and should be classified as suspicious/high-risk.