expression-language-injection

SUSPICIOUS/HIGH RISK. 该 skill 与声明目的一致，但其目的本身就是为 AI 代理提供 EL 注入与 RCE 利用方法，包括文件读取、命令执行、外带和 CVE 攻击。没有安装器、凭证窃取或隐藏执行证据，因此不属确认恶意软件；但作为进攻性安全技能，对目标系统具有高现实危害。

The fragment is exploit-oriented documentation containing working-looking payloads for expression-language injection, command execution, data exfiltration, information disclosure, and authorization/session manipulation. It should not be evaluated with attacker-controlled inputs. As shown, it is textual content rather than executable package logic, so there is no direct evidence of malware behavior or an actual network connection; the risk would become high if surrounding code evaluates or dispatches these examples.