Skill Ship Safety

Audits static safety evidence for a skill that sends email / WhatsApp or triggers other external actions (安全真发): verifies documented commands exist, identifies external-action code, and records whether trusted isolated behavior tests remain unverified. Returns a static pass / stop / unverified result without executing target code. Use when the user asks 安全真发审计, 能不能真发, ship-safety review, or wants an intent-vs-implementation check before the first real send.

xenos2025 Updated
模型调用拓扑与功能角色拓扑

File contents

xenos2025/skill-self-check commit 072525829a

Frequently asked questions

npx get-skill add xenos2025/skill-ship-safety