websocket-attack

SUSPICIOUS/HIGH-RISK skill. Its footprint is coherent with its stated purpose, but that purpose is to equip an AI agent with offensive security techniques, including injection, token theft scenarios, localhost probing, and exploit validation. The added clone-and-run third-party tool instructions raise supply-chain risk, especially for unpinned/personal repos. Not confirmed malware, but dangerous and inappropriate for broad agent use.

该输入不是可执行 npm 依赖模块，而是高风险的 WebSocket 攻击与漏洞利用指南。其示例明确包含跨站会话劫持、敏感数据外传、本地端口扫描、潜在远程命令执行、代理隧道、DoS 和原型链污染技术。若作为软件包内容发布或自动运行，应视为严重安全风险；仅作为受控授权测试文档时，仍需严格限制使用范围。

This fragment is not malicious executable code; it is a set of structured, attack-oriented prompts meant to generate actionable guidance for exploiting WebSocket/Socket.IO systems (including handshake vs. message authorization gaps, protocol framing, and passive monitoring). From a supply-chain perspective, the primary risk is misuse/weaponization of the artifact rather than runtime malware behavior.