Stop AI from burning tokens and over-analyzing. 防止 AI 插件乱烧 Token 和过度分析。
AI plugins (like Superpowers) are often "too diligent"—re-reading the entire project for a simple fix or re-planning everything after a short break. This Skill gives the control back to you.
AI 插件往往“过度勤快”:修个小错别字也要遍历全项目,或者中断回来就要重读所有文档。本协议通过逻辑门控,把插件的“开关”重新握在用户手中。
AI must ask for permission before performing heavy scans or global searches. AI 执行重型扫描或全局搜索前必须先“请示”。
Skip re-reading docs by using a 3-line [Session_Anchor].
通过 3 行“会话锚点”跳过文档重读,直接进入开发/任务状态。
Force AI to focus only on the reported error or the current file. 强制 AI 仅关注报错区域或当前文件,严禁“想太多”。
- Copy (复制): Copy the content of
config/gatekeeper.json. - Inject (注入): Paste it to your AI (Claude, Cursor, GPT-4) and say:
"Load this skill and follow the logic gate rules. / 加载此技能并遵守逻辑门控规则。"
- Control (控制):
- Choose Local-Only to save tokens. (选“本地处理”来节省 Token)
- Say "Archive / 存档" before you leave. (离开前说“存档”)
- Paste the anchor and say "Continue / 继续" when you are back. (回来后贴上锚点说“继续”)
/config/gatekeeper.json: Machine-readable logic. (机器可读逻辑)/docs/guide_en.md: Detailed English guide./docs/guide_zh.md: 详细中文指南。
MIT License. Created for the Vibecoding community.
备注:我是因为修复 Bug 和断点重连太费 Token 才想做这个,实战中可以这样玩:
当 AI 弹出那句“
Note: I’m doing this because fixing bugs and re-establishing breakpoints wastes too many tokens.This trick works in actual use:
When the AI pops up the message:“