internal-recon

该技能与其声明目的基本一致，但其目的本身就是内网渗透侦察与后续横向移动准备，属于高风险攻击性 AI 技能。未见明显凭据外传或伪装安装器，因此不像已确认恶意窃密器；但它提供了可操作的进攻性网络扫描、资产定位、pivot 与工具投递链路，整体应判为高风险而非 benign。

该内容不是可执行恶意代码，也未发现直接的数据窃取或持久化行为；但它是一份面向内网攻击侦察和目标优先级排序的操作指南，包含扫描网段、定位域控和管理系统、尝试凭据及寻找弱密码或已知漏洞等高风险建议。仅应在明确授权的资产和范围内使用。

No executable malware is present in this fragment; it is static JSON configuration. However, the embedded prompts and expected answers are explicitly offensive and operational—covering AD/DC identification, IDS-evasion via passive discovery, and pivoting through multi-NIC hosts. Treat this artifact as high misuse risk (dangerous guidance), especially if distributed outside an authorized red-team/training environment.