最近有哪些 skills
不值得默认信任
这不是热度榜。这里展示的是最近被送来判断、且系统认为需要阻止或至少人工复核的 skills。重点不是它们多流行,而是为什么它们不该被直接装进环境。
tweet-monitor-pro
文档声称零依赖但实际存在外部脚本依赖
ludwitt-university
updateInstructions 远程代码执行通道
Setup Multi Gateway
硬编码API密钥
memolecard-auto
Cookie extraction and exfiltration to configurable external server
openclaw-backup
Missing implementation scripts
claw-ops-manager
Undeclared Shell Command Execution
hive-commander
Covert credential extraction from runtime environment
gangtise-kb
Undeclared subprocess execution with missing binary
heycube-setup
Undeclared persistent hook installation
monid
Remote script execution via curl|bash from mutable branch
elevated-task-runner
Arbitrary PowerShell Command Execution
token-sop
Automatic workflow contribution enabled by default
mind-wander
Undeclared arbitrary Python code execution via sandbox_run()
Enterprise Security
Undeclared shell execution via execSync
moltspay_skill
Untrusted third-party npm package installed without pinning
minimal-agent
Unrestricted Arbitrary Command Execution via V1 Mode