AI Decision System
Replayable, offline-verifiable records for AI-assisted go/no-go decisions: experiment promotions, typed model answers and human-approved actions.
翻译准备中
个人 GitHub 账号下公开仓库的所有者可以登录并验证所有权,自行管理介绍,无需管理员批准。
登录并生成验证文件,在 30 分钟内将其提交到此仓库的默认分支。其他分支或分叉仓库中的文件无效。然后在下方检查文件,即可无需管理员批准直接管理项目。请保留文件以维持管理权限。
如需登录,将前往登录页面。登录后请回到此处,再次点击所选的验证按钮。
在项目文件夹中运行 /nomorevibe verify,即可验证所有权并自行管理介绍。
一分钟完成验证Open-source software factory: one lead per project, fleets of coding agents across your machines
翻译准备中
取自仓库 README
正在加载完整 README。
查看完整 README ↗fleetd runs Claude Code on your machines: you type what you want, and it plans the work, runs the agents, opens pull requests it has checked, and asks you only about real decisions.
curl -fsSL https://fleetd.sh | sh
You need: macOS or Linux x64, git, and the claude CLI where agents run. Claude runs on the hub's API key, your
subscription or each machine's login; GitHub through the hub's own GitHub App, or gh. fleetd init walks you
through them, the same steps as the app's first run. About a minute later the hub runs on this machine, the app is
paired, and you send the lead your first message.
Docs: fleetd.sh/docs Contributing: docs/CONTRIBUTING.md License: MIT