race-condition-methodology

Fail

Audited by Socket on Mar 30, 2026

3 alerts found:

SecurityMalwarex2
SecurityMEDIUM
evals/evals.json

No malware execution exists in this snippet, but it is explicitly adversarial instructional material that facilitates TOCTOU/race-condition exploitation for fraud and webshell access (including concrete targets like POST /api/buy and /uploads/shell.php). Any inclusion in a benign project should be treated as a serious misuse risk and should be removed/isolated unless there is a clear, controlled defensive/academic justification.

Confidence: 80%Severity: 90%
MalwareHIGH
references/race-scripts.md

This code is intentionally malicious/offensive: it automates exploitation of race conditions to cause duplicate state changes (double-spend/duplicate redemption), deploys a PHP webshell enabling remote command execution, and automates exfiltration of sensitive files. Treat as exploit tooling/backdoor; do not run against systems unless under an explicit, authorized engagement. If found in a codebase unexpectedly, remove and investigate source of inclusion and any credential exposure.

Confidence: 75%Severity: 95%
MalwareHIGH
SKILL.md

该技能不是普通开发辅助,而是面向 AI 代理的漏洞利用指南,直接教授并自动化条件竞争攻击,对目标系统可造成重复消费、库存绕过、token 重放和上传竞态利用。虽无明显供应链或隐蔽窃密链路,但其能力与“exploit”定位一致且高风险,应判定为高危攻击型技能。

Confidence: 95%Severity: 91%
Audit Metadata
Analyzed At
Mar 30, 2026, 02:22 AM
Package URL
pkg:socket/skills-sh/wgpsec%2Faboutsecurity%2Frace-condition-methodology%2F@1167d75ef2da7ea7f1643e59267800a80822162d