A few months ago I was 40 minutes into a retry spiral on a bug that turned out to be three lines. Not because the bug was hard because it was 1am, I'd been at it for six hours straight, and I genuinely could not tell anymore that the thing that was broken was me, not the code.
I built Tempo for that specific moment. It's a free, local-first companion that reads the shape of your Claude session retry loops, hours logged, that particular late-night urgency and never the content of your prompts, code, or files. When the pattern says you've checked out, it offers exactly one thing: a 40-second reset, at a natural pause, never mid-task. Max two offers a day.
The question I get most from other devs when I describe Tempo is some version of "wait, so it's reading my code?" No and that was a hard constraint from day one, not a privacy-policy afterthought.
Tempo only ever looks at session metadata: how long you've been active, retry/error frequency, time of day, gaps and bursts in activity. It never touches prompt content, file contents, or output. There's no account to create and no network calls in the codebase at all you can check this yourself, it's not a claim you have to take on faith.
Every break-reminder feature I've tried eventually gets snoozed into oblivion mine included, in early builds. Curious what's actually worked for people, or what's made you turn a feature like this off entirely.
A load-aware companion for Claude — chat, Cowork, Claude in Chrome, and Claude Code. Tracks how long you have actually been in Claude, on from minute one, fully local, never interrupts your work.
