prototype-pollution-exploit

Fail

Audited by Socket on Apr 22, 2026

2 alerts found:

SecurityMalware
SecurityMEDIUM
SKILL.md

该技能不是凭证窃取或供应链投毒型内容,但它的核心功能是教 AI 代理实施原型链污染攻击并达成 RCE/XSS/权限绕过,属于高风险 offensive security skill。安装与数据流相对干净,恶意意图未达确认恶意软件级别,但安全风险应判为高。

Confidence: 94%Severity: 84%
MalwareHIGH
references/server-side-rce.md

The provided code is highly indicative of malicious intent: it is an exploit write-up/PoC showing how to use server-side prototype pollution (__proto__) to achieve RCE through multiple Node.js template engines and child_process option poisoning, including explicit out-of-band exfiltration examples. If this were present in a package on a public registry, it would be an extremely dangerous supply-chain artifact. Note: the snippet appears to be offensive tooling/instructions rather than normal dependency code, but as submitted, the security risk is still very high.

Confidence: 90%Severity: 95%
Audit Metadata
Analyzed At
Apr 22, 2026, 08:00 AM
Package URL
pkg:socket/skills-sh/wgpsec%2FAboutSecurity%2Fprototype-pollution-exploit%2F@17445766daae81d6a58c848d4520095213c48f0c