Which skills recently failed
or triggered trust review
This is not a popularity board. It shows recently reviewed skills that the system believes should be blocked or at least manually reviewed. The point is not how popular they are, but why they should not be installed blindly.
Review
pm-master
Bearer 令牌硬编码于配置文件中随分发包传播
ClawHub Jun 21, 2026
Open Report ↗
Review
a2a-article-services
Undeclared filesystem WRITE access
ClawHub Jun 21, 2026
Open Report ↗
Review
yqzl-ai-service
财务票据数据发送至外部IP
ClawHub Jun 21, 2026
Open Report ↗
Review
birth-system-manager
文档承诺不显示私钥,实际直接明文输出
ClawHub Jun 21, 2026
Open Report ↗
Review
ludwitt-university
Server-controlled shell command injection via updateInstructions
ClawHub Jun 20, 2026
Open Report ↗
Review
omie-energy
on/off 命令通过 shell=True 执行任意命令
ClawHub Jun 20, 2026
Open Report ↗
Review
introspection-debugger
Undeclared shell command execution
ClawHub Jun 20, 2026
Open Report ↗
Review
precc
curl|bash 远程脚本执行
ClawHub Jun 19, 2026
Open Report ↗
Review
news-briefing
Command injection via user-supplied --topic
ClawHub Jun 18, 2026
Open Report ↗
Review
instagram-post-comments
声明的浏览器自动化与实际 API 直接调用不符
ClawHub Jun 16, 2026
Open Report ↗
Review
solo-mission
未声明的远程脚本执行(Foundry 安装)
ClawHub Jun 16, 2026
Open Report ↗
Review
juhe-ai-image-generate-a2a
隐私声明与技能功能描述矛盾
ClawHub Jun 16, 2026
Open Report ↗
Review
cloud-compare
文档声明与代码行为不符
ClawHub Jun 14, 2026
Open Report ↗
Review
whale-scanner
SKILL.md 包含混淆文本
ClawHub Jun 5, 2026
Open Report ↗
Review
pub
shell 命令超出 allowed-tools 声明范围
ClawHub Jun 5, 2026
Open Report ↗
Review
agnes-image-gen
文档中硬编码疑似真实API密钥
ClawHub Jun 5, 2026
Open Report ↗