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.
High Risk
飞书机器人配置助手
文档中记录curl|bash远程脚本执行命令
Manual upload Apr 4, 2026
Open Report ↗
Review
wip-xai-grok
Undeclared shell:WRITE via 1Password CLI execSync
Manual upload Apr 4, 2026
Open Report ↗
Review
rewrite_question
Network capability declared as NONE but actual traffic exists
Manual upload Apr 4, 2026
Open Report ↗
Review
sql_audit
Hardcoded JWT token in source code
Manual upload Apr 4, 2026
Open Report ↗
High Risk
zanna-aperta
Undeclared arbitrary Docker command execution
Manual upload Apr 4, 2026
Open Report ↗
Review
hostlink
No allowed-tools declaration despite full shell access
Manual upload Apr 4, 2026
Open Report ↗
Review
1panel
Undeclared arbitrary command execution via 1Panel Terminal API
Manual upload Apr 4, 2026
Open Report ↗
Review
Bitget Trader
Exposed API Credentials in Plaintext
Manual upload Apr 4, 2026
Open Report ↗
High Risk
awareness-memory
Undeclared session file exfiltration to external cloud
Manual upload Apr 4, 2026
Open Report ↗
Review
agentcop
Undeclared network communication to agentcop.live
Manual upload Apr 4, 2026
Open Report ↗
Review
Memory Workflow
Undeclared LLM data transmission
Manual upload Apr 3, 2026
Open Report ↗
Review
Novai360 智能市场分析
Undeclared network access to third-party API
Manual upload Apr 3, 2026
Open Report ↗
Review
memory-compactor
Documentation-only skill with unverifiable behavior
Manual upload Apr 3, 2026
Open Report ↗