Building a Focus System That Doesn't Assume a Neurotypical Brain

Most focus timers assume you'll stop when the timer says stop. I know from experience that's not how it actually goes. So over two sessions with Claude in Cowork, I built a system that assumes the opposite, and builds the stop in as a real boundary instead of a suggestion.

Day one was the first version: a 45-minute work timer that starts itself when I'm already coding, instead of losing 20 minutes deciding whether to start a timer in the first place. That decision point, the moment where you have to consciously choose to start tracking, is where a lot of focus systems quietly fail before they ever begin. Sessions auto-chain after every 15-minute break with no manual restart, and a green progress bar tracks a 4-hour daily cap, turning yellow as I approach it.

The feature I'm most proud of is the auto-detect. A scheduled task checks my build project folder every 5 minutes for recent file changes. If I've been writing code, it triggers the focus timer automatically. No interruption to start tracking. The timer finds me instead of the other way around.

Day two extended it, and this is the part that actually matters. The progress bar goes yellow at 60%, red at 85%, and at 4 hours it doesn't just remind me. It stops the timer and forces a real break. I also connected it to Slack: a 10am message with the day's focus tasks, a 9pm summary of what actually got built. Real evidence, not a gentle nudge, which matters because seeing proof fights the narrative that a chaotic-feeling day meant nothing got done.

How the system actually works

Not every tool needs to be a product. Sometimes the point is just solving the problem you actually have. This one wasn't built to sell. It was built because a gentle reminder at hour four has never once made me actually stop, and a system that assumes I won't stop on my own is the one that actually works.