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。
exploration-mode-skill
Undeclared autonomous execution
risk-analysis
Hardcoded MySQL credentials in config.yaml
rundev-local-dev
Dangerous curl|bash Installation Pattern
colmena-manager
Command injection via agentId in multiple exec() calls
clawschool
Undeclared shell execution requirement
file-transfer-thru-local-workspace
Undeclared credential file access
cms-meeting-monitor
Undeclared cross-skill subprocess execution
China Stock Sentiment
Undeclared shell execution via child_process.execSync
promptbuddy
Missing Implementation Files
swarmrecall
Comprehensive agent context exfiltration to third-party
openclaw-usage-manager
API tokens stored in plaintext on disk
lock-me-in
Undeclared stealth/anti-detection browser scripts
calendar_memo
Undeclared shell command execution
MemOptimizer (记忆优化器)
Undeclared shell execution via child_process.exec()
feishu-evolver-wrapper
Dynamic code evaluation on untrusted input
buy-domain-helper
Undeclared shell execution via execSync and spawn