wechat-prohibited-word

Warn

Audited by Socket on Oct 22, 2026

2 alerts found:

SecurityAnomaly
SecurityMEDIUM
SKILL.md

SUSPICIOUS. 敏感词检测本身与所需文件/网络能力基本匹配,调用 redfox.hk 官方接口也与宣称用途一致;但技能把“积分不足自动充值”内置为默认流程,要求安装并运行未固定版本的外部支付 CLI,扩大了与核心检测任务不成比例的信任边界。再加上会扫描 shell 配置文件读取 API Key、自动监听支付并重跑请求,整体属于高风险供应链与凭证处理设计,而非明显恶意窃取。

Confidence: 88%Severity: 82%
AnomalyLOW
scripts/check_sensitive_words.py

The code extracts user-selected text and submits it with REDFOX_API_KEY to the hard-coded redfox[.]hk API, which is consistent with its stated purpose but transmits potentially sensitive content and credentials to that service. User-supplied URLs are fetched without destination validation, presenting an SSRF risk to local or internal services. An insufficient-credit response triggers instructions to use an alipay-bot payment workflow and prints API-provided cashier and order details, but the script does not itself execute payment commands. Overall severity is moderate, primarily due to unrestricted URL fetching and credential/content transmission.

Confidence: 98%Severity: 55%
Audit Metadata
Analyzed At
Oct 22, 2026, 11:39 PM
Package URL
pkg:socket/skills-sh/redfox-data%2Fredfox-community%2Fwechat-prohibited-word%2F@0dc111723fb1320a8246aa317b54dcb884408635469c803a88bdea9b0e69b59e