A time-and-motion study of your computer and you

Efficiency had a stopwatch. Then a camera. This time, you study yourself. Gilbreth quietly records your real workday: app switches, windows, input rhythm, idle time. Every byte stays on your machine, and the dashboard shows you the wasted motions worth removing.

Download for Windowsv0.1.2 preview • Windows 11 (x64) Read the source

Free & open source. Windows SmartScreen may warn while the signing certificate is new — verify the download yourself.

Frank Gilbreth observing a worker at a factory machine, photographed during a motion study.

Screenshots frame by frame

Gilbreth's Analytics tab: active-time and keystroke headline figures above a day-by-hour heatmap of working rhythms. 01 Gilbreth's Analytics tab: Patterns Worth Reviewing cards above focus metrics and interruption-cost figures. 02 Gilbreth's Week tab: a weekly active-hours heatmap, per-app time bars, the morning launch sequence, and patterns that changed this week. 03 Gilbreth's Recordings tab: a recorded routine with its replay-readiness verdict, step table, and value-free export options. 04 Gilbreth's Privacy tab: the live data posture, redaction and exclusion rules, retention settings, and erase controls. 05

Tap a frame to see it full size.

How it works

Study the motions. Remove the friction.

01

Use

App captures actions locally

02

Inspect

Visualize your real workday

03

Find

The wasted motions, surfaced

04

Plan

Develop personal automation

Use

A small, quiet app records a structured motion trace: app switches, windows opening and closing, keystroke timing and classes, clicks, pointer motion, and idle time. Events are timed to the millisecond and stored in a local database on your machine; privacy controls and noise filters deliberately bound what persists.

The Gilbreth tray menu with the Capture submenu open: checkmarks beside the Foreground, Windows, Keyboard, Mouse, System, and Idle streams.

Inspect

A dashboard turns the record into a picture of your actual workday: where your time pooled, what you switched to, which apps always travel together.

The Rhythms heatmap: a week of active hours laid out by day and hour, brighter where more active.

Find

Sift the record for repeated rituals and friction to discover the digital wasted motions. Pattern-finding is heuristic today; ML-driven suggestions are next on the roadmap (see where we are, below).

Two Patterns Worth Reviewing cards: a long-stretch input-exposure pattern and a short-lived-windows routine, each with signal and recurrence counts.

Plan

Findings become specific suggestions, like an app that does your ritual in one step, a custom AI tool shaped to your exact workflow, or an integration that makes two stubborn apps finally talk. These are your motions, your fixes.

The Recordings export panel: buttons to export a value-free trace or copy the analysis prompt, above the prompt text that ships with it.

Privacy

Watched by no one. Studied by you.

A motion study needs a broad, consistent record. Gilbreth captures app/window activity, input timing and motion, and system context; typed key content stays off unless you opt in.

Local only, by design

Everything lives on your machine — your activity in one database file, plus a couple of small settings files beside it. The native dashboard opens no network listener, and the app makes no outbound network calls. There is no server to send your data to, no telemetry, no cloud.

You hold every switch

Pause all ambient capture with Ctrl+Alt+Shift+P or the tray's Pause capture item. The tray icon visibly changes while paused, and capture resumes only when you press the chord again or choose Resume capture.

You can see and destroy everything

Inspect every recorded row. Delete any of it. Or erase it all, securely, in one action. Old history can prune itself on a retention schedule you set, and erasing means a true secure wipe.

Where we are

The first release is out

A week of normal use is enough for Gilbreth to show you where your time pooled, what a pull-away costs you on return, and which rituals repeat often enough to be worth automating.

Gilbreth v0.1.2 is on GitHub as an early preview: a signed installer for Windows 11 (x64). Features include ambient capture, a privacy filter, local storage, and a native dashboard with seven tabs, Today through Diagnostics. The pattern-finding is heuristic today: routine motifs, fragmentation and interruption costs, input exposure. ML-driven suggestions and a packaged macOS build are next on the roadmap.

The source is public under AGPL-3.0, and you don't have to take the privacy claims on faith: verify the download before you run it.

Download for Windows

Setup takes about a minute. Gilbreth lives in the system tray, the dashboard opens from the tray menu, and the patterns get interesting after a few days of normal use.

macOS capture already works in development builds. On a Mac? Join the release list below and you'll hear when it ships.

One email per release worth announcing. Nothing is tracked, and replying to any email unsubscribes you.

The Gilbreth dashboard's Today tab: a timeline of which app was in front, an hourly input pulse, active-time and focus-switch figures, and patterns flagged as worth reviewing.
The dashboard's Today tab, rendered from the release's own test-fixture data.

FAQ

Questions about Gilbreth

What's the point of Gilbreth?
Most of us work all day inside habits nobody has ever looked at. Gilbreth records how you actually use your computer, then shows you the patterns: where time pools, what interrupts you, which little rituals you repeat hundreds of times. Some of that is fine. Some of it is worth automating or dropping, and you can't decide until you've seen it.
What can I expect to learn?
Within a day: where your time went, hour by hour, and how much was active work versus idle. Within a week: your working rhythms, what pulls you away and what it costs to come back, and which app-to-app routines repeat enough to matter. It will also tell you when you work in long stretches without a break.
Is my data portable?
Yes. Everything lives in one SQLite database on your machine, and SQLite is about as portable as data gets: point any tool you like at it. The tray can export a passphrase-protected archive (or a plain copy of the database, if you accept the risk), and recorded routines export as plain JSON you can hand to any AI or automation tool. No account, no cloud: the file is already yours.
Is my data file safe?
As safe as the account it lives in. The database sits in your Windows user profile, so ordinary file permissions keep other accounts out, and there is no cloud copy for anyone to breach. It is not separately encrypted while live: anyone who can unlock your Windows session can open it, the same as your documents and browser history. Archives are encrypted, and portable exports take a passphrase.
Is this a time tracker?
No — time trackers are stopwatches. They tell you how long. Gilbreth is more like a camera: it shows you which motions you do the most, so the waste can actually be removed.
Is this spyware / bossware?
No. Bossware sends your activity to someone else. Gilbreth makes no outbound network calls at all — your data never leaves your machine, and you're the only person who can see it. The source is public, so you don't have to take that on trust.
Does it record keystrokes?
It can — a motion study with the keyboard missing isn't much of a study. By default it counts your keystrokes and their timing, never which keys you pressed; storing actual key content is a separate, explicit opt-in. Every stream can be switched off, and you can inspect or erase everything it stores.
How is this different from ActivityWatch or RescueTime?
ActivityWatch is local and open source too; it records your activity and hands you the charts. RescueTime is a cloud service with accounts and team dashboards. Gilbreth is built for the step after capture: finding the repeated friction (routines, fragmentation, input strain) worth removing, entirely on your machine.
How do I uninstall?
The normal Windows way: Settings, Apps, uninstall. Your recorded data stays on disk by default so you can come back later. If you want it gone too, use the tray's Erase all my data before uninstalling and Gilbreth does a secure wipe. Nothing is left running and no autostart entries remain.
What's Gilbreth built with?
Rust, end to end: the capture engine, the storage layer (SQLite), and a native dashboard built on egui (no web view, no bundled browser). One executable plays both roles. The source is on GitHub under AGPL-3.0 if you want to see how any of it works.
Where does the name come from?
From Frank and Lillian Gilbreth, who invented motion study a century ago: they filmed work, found the wasted motions, and made jobs less exhausting. Lillian's rule was that the study serves the worker, and this project tries to keep that rule. The full story is below.
When can I use it?
Now, as an early preview. v0.1.2 is on GitHub: a signed installer for Windows 11 (x64) you can verify before running, or build it from source. macOS capture works in development builds; a packaged macOS release comes later.

Why it matters

Your work has its own wasted motions.
You've just never seen them.

A century after the Gilbreths, work moved onto screens, and went invisible again. No film to study or hands to watch.

The software that claims to help mostly brought back the stopwatch: hours-in-app charts and screen-time guilt. Its darker cousin, bossware, turned motion study into something done to workers, for someone else's benefit.

Gilbreth is the other tradition, offering efficiency that serves the worker. It's the camera, handed back to the person doing the work.

Gilbreth micromotion film still: a typist at her machine with the Gilbreth microchronometer clock in frame.

The story

The Gilbreth story:
Finding the one best way

In 1885, a young bricklayer named Frank Gilbreth noticed something strange: no two masons laid brick the same way, and none of them could say why they moved the way they did.

So he started watching. Then filming. Frank and his wife Lillian, one of the first industrial psychologists (later called the First Lady of Engineering), broke work down into its elemental motions: search, find, grasp, position, rest, plan. They called these units therbligs (their name spelt backwards, roughly). They strapped small lights to workers' hands, photographed long exposures, turned invisible habits into visible ribbons of light, and called it the chronocyclegraph. For the first time, anyone could see wasted motion.

The results were stunning. Bricklaying went from eighteen motions per brick to about four and a half. Output nearly tripled — and the masons went home less tired.

That last part matters. While Frederick Taylor stood over workers with a stopwatch demanding speed, the Gilbreths removed the motions that wasted effort. Lillian's rule was non-negotiable: efficiency serves the worker, never the other way around.

(You may already know this family — two of their twelve kids wrote Cheaper by the Dozen about growing up inside the experiment.)

Portraits of Lillian and Frank Gilbreth. A Gilbreth chronocyclegraph photograph: ribbons of light tracing a worker's hand motions beside the Gilbreth timing clock.

Find your one best way.

The Gilbreths proved a century ago that the kindest thing you can do for work is to see it clearly. It's your motion. Study it.

Download for Windows Join the release list