hping3

The skill is internally consistent with its stated purpose and does not show credential theft or suspicious install behavior, but it equips an AI agent with privileged packet-crafting and penetration-testing functionality. That makes it high security risk as an agent skill even though malicious intent is not evident.