轻量待办,自由计时
Lightweight tasks, flexible timing
写下今天要做的事,开始工作时点一下。不规定 25 分钟。
Write down today’s tasks and start with one tap. No forced 25-minute sessions.
你的工作已经多线程了,计时器也应该是。
Your work is multithreaded. Your timer should be too.
轻量待办 + 自由计时。多个任务可以同时运行,重叠的真实时间不会重复计算。不规定 25 分钟,只记录今天真正做了什么。
A lightweight daily task list with flexible timing. Run multiple tasks at once without double-counting overlapping time. No forced 25-minute sessions.
时流只保留真正有用的部分:开始、暂停、完成、回看。
ThreadFlow keeps only what matters: start, pause, finish, review.
写下今天要做的事,开始工作时点一下。不规定 25 分钟。
Write down today’s tasks and start with one tap. No forced 25-minute sessions.
AI Agent、代码构建或内容生成在跑时,可以自然切到另一项工作。重叠的真实时间只计算一次。
Switch naturally while AI agents, builds, or generation tasks run. Overlapping real time is counted only once.
不做复杂项目管理。只留下今天做过什么,以及真正投入多久。
No complex project management. Just what you did and how long it really took.
查看最近 7 天和每月工作时间统计
Review 7-day and monthly work time statistics.

随手写下要做的事,开始工作时点一下。Add what you want to do and start with one tap.

回看每天做过的事,以及每项工作的实际投入。Look back at what you did and how much time it took.

看看时间都花在了哪里。See where your time went.
ThreadFlow 不规定 25 分钟的工作周期。开始工作时开始计时,需要暂停时随时暂停,更适合不希望被固定专注周期打断的人。ThreadFlow does not require 25-minute work sessions. Start and pause when you need to, without a fixed focus cycle interrupting your work.
可以。ThreadFlow 支持多个任务同时运行。当任务发生时间重叠时,真实经过时间会自动分配,因此总工作时间不会被重复计算。Yes. When tasks overlap, real elapsed time is automatically allocated, so your total work time is never double-counted.
适合。当 Codex、Claude Code 或其他 AI Agent 正在执行任务时,你可以切换到另一项工作,并同时保留多个任务的运行记录。Yes. While Codex, Claude Code, or another AI agent runs, you can switch to another task and keep both work records running.
不会。当前版本无需登录,任务、分类和工作时间记录均保存在设备本地。No. The current version does not require sign-in; tasks, categories, and work-time records stay on your device.
ThreadFlow 是一个开源项目。
欢迎提交 Issue、分享想法、报告问题,也欢迎通过 Pull Request 一起完善产品。
ThreadFlow is open source.
Ideas, bug reports, feedback, and pull requests are all welcome.
多任务可以并行,时间不会重复计算。
Run tasks in parallel without double-counting your time.