我在5天内通过Taskwarrior和Claude Code完成了706次提交。
网站:https://ttal.guion.io
706次提交,38个合并的PR,5个代码库,5天。一个人最多进行5个Claude Code会话。
技术栈:Taskwarrior作为任务队列,Zellij作为会话管理器,Claude Code作为工作者。每个CC会话在一个Zellij窗格中运行,并分配一个任务。当一个会话完成时,下一个优先级最高的任务会通过Taskwarrior的钩子自动启动。你管理的是任务,而不是会话。
周四我遇到了API速率限制,吞吐量下降了75%。这就是证据——系统是瓶颈,而不是我。
关键设计:按需的人机协作。代理从不因等待我而阻塞。它们接手任务,完成工作,提交,然后继续前进。我在准备好时审查PR并做出决策——而不是在代理需要我时。这就是消除人作为瓶颈问题的方法。当没有会话在等待你时,5个会话就足够了。
完整的设置已在网站上记录。架构并不局限于Claude Code——Zellij会话不在乎里面运行的是哪个CLI代理。
查看原文
Site: https://ttal.guion.io<p>706 commits, 38 PRs merged, 5 repos, 5 days. One person with max 5 Claude Code sessions.<p>The stack: Taskwarrior as the task queue, Zellij as the session manager, Claude Code as the worker. Each CC session runs in a Zellij pane with a task assigned. When a session finishes, the next highest-urgency task auto-starts via Taskwarrior hooks. You manage tasks, not sessions.<p>Thursday I got API rate-limited and throughput dropped 75%. That's the proof - the system was the bottleneck, not me.<p>The key design: on-demand human-in-the-loop. Agents never block waiting for me. They pick up tasks, do the work, commit, and move on. I review PRs and make decisions when I'm ready - not when the agent needs me. That's what eliminates the human-as-bottleneck problem. 5 sessions is plenty when none of them are waiting on you.<p>The full setup is documented on the site. Architecture isn't locked to Claude Code - Zellij sessions don't care what CLI agent runs inside them.