Log in Get started

For principals

Classroom apps that keep school data in the school.Structure for every student, data that stays in the school.

Each teacher runs the class server on a computer the school owns. Rosters, problem sets and answers never leave it for a vendor's cloud.Every assignment arrives as a sized, dated card with one at the top, the kind of visible structure that helps students with ADHD keep track. Each teacher runs the class server on a computer the school owns.

Keys and pricing Log in

Not you? Pick again.

keeplaunch
  • keeptrack
  • keepcourses
  • keepinstruct
  • keepsetup-instruct
  • keeptraining
  • keepconverting
  • keeplaunch

What you get

Made for principals.

Your school's data stays in the school

The class server is a program on a computer you own. Nothing is uploaded to us or to anyone else.

Encryption built for the next decade

Every connection uses the new post-quantum standards, so what is recorded today cannot be read later by a stronger computer.

Keys you can count and revoke

A class key covers one class and a set number of people. A paid key is one person. Every key is listed and can be revoked.

Windows, macOS and Linux, no accounts elsewhere

Students install from the copy the server publishes. No GitHub account, no cloud sign-up, no analytics on this site.

ADHD and these apps

Built around what the research says helps.

Every feature below is in the apps today. Each carries a number that links to the study behind it, listed in full under works cited at the foot of the page.

Size, due date and progress on every card

Barkley's account of ADHD puts working memory and the sense of time at the centre: holding a plan in your head and feeling how far away a deadline is are the hard parts. His practical conclusion is to put that information outside the head, where it can be seen. keeptrack does exactly that: each assignment is a card with its size (ω), its due date and a progress slider you can see at a glance. [1]

One card at the top

The dashboard sorts by urgency, so there is always one card to start on. Deciding in advance what you will do next is what the research calls an implementation intention, and a meta-analysis of 94 studies found it has a medium-to-large effect on actually following through (d = 0.65). A later experiment found the same if-then planning improved response inhibition in children with ADHD. [2] [3]

Progress you move, and a score that moves with it

Monitoring progress toward a goal makes reaching it more likely, and more so when the progress is recorded somewhere rather than only noticed, according to a meta-analysis of 138 studies. The slider on each card is that record; the σ score and the late toast are the feedback. [4]

Time estimates in work, not in hours

People underestimate how long their own tasks will take, the planning fallacy, and adults with ADHD reproduce and estimate time intervals less accurately than controls. keeptrack asks for a size in ω instead of a guess in hours, then shows urgency as work left against time left, so being behind pace is visible before the deadline is. [5] [6]

Work broken into sets, kept on the screen

Working memory is one of the most consistent differences found in ADHD across a meta-analysis of 26 studies. keepcourses keeps each course's readings, problem sets and progress in one folder, keepinstruct hands out problems one set at a time, and the editor keeps the file and the card side by side, so less has to be held in mind. [7]

Reminders that cost something

Each time keeptrack opens, every past-due, unfinished assignment costs a little score and a toast says which one. That is a reminder tied to a consequence, at the moment you can act on it; Barkley's work argues that for ADHD, cues and consequences work best at the point of performance, not in a lecture beforehand. [1]

These apps are a study aid. They do not diagnose or treat ADHD or anything else, and the studies above are about planning, memory and time, not about these apps.

The apps

Everything keeplaunch can open.

Each app is its own window with its own job. Hover or tap one to read what it does.

  • keeptrack
    keeptrack

    Assignments and deadlines, with progress estimates, urgency and a score. Courses, readings, mail and AI help in one window.

  • keepcourses
    keepcourses

    Every course in one place: syllabi, readings, assignments and progress, plus an optional ADHD homework upgrade that rewrites trivial problems into harder, more engaging ones.

  • keepinstruct
    keepinstruct

    Teach a class: drop problems on your students live and watch their answers arrive.

  • keepsetup-instruct
    keepsetup-instruct

    Your students' setup codes, the class keys for your classes, and who is waiting for you to approve them.

  • keeptraining
    keeptraining

    Practice sessions and drills, quizzes and tests, with stats for the last 14 days.

  • keepconverting
    keepconverting

    Turns a Markdown file into a PDF or a Word document, in MLA format if you want.

  • keeplaunch
    keeplaunch

    The launcher. Every keep app on one grid, each a click away.

Hover or tap a tile to read what it does.

Security

Your class data stays on your own computer.

The class server runs on a computer the host owns, every connection to it is post-quantum encrypted, and the models that estimate your progress run on your own machine.

The four protections, in detail

Post-quantum encryption

Every connection to the class server uses the new post-quantum standards (ML-KEM and ML-DSA, the standardised CRYSTALS-Kyber and Dilithium). A client that can't do it is refused, never downgraded.

Self-hosted

The class server is a program on the host's own computer. Rosters, problem sets, answers and the published copy of the suite stay there.

Local AI

Laya and Gemma run on your own machine. Progress estimates come from the file you are editing, with no call to a cloud model. The optional Ask AI and Sources features can send your question to Claude through a sign-in you make yourself.

Signed keys and one-time codes

Every licence key is signed, so it cannot be forged without the owner's private key. A setup code works once, to choose a password of at least 12 characters; after that the password alone signs you in.

Pricing

One key per teacher, one class key per class.

Each teacher needs their own key on one of the plans below. Institution keys are how a school gets keys for its instructors: the owner issues them to you and you hand them out. Students come in on class keys, one per class, sized 1 to 1000.

Yearly

$3 an app

$3 an app and $2 a year for the first year, then $2 a year in all, however many apps.

One app for one year is $5. The $2 a year keeps the key server running.

One person, one role.

Buy it

Lifetime

$8 an app

$8 an app, once. No renewal.

One person, one role.

Buy it

Ultimate

$75 once

Every paid keep app there is now and every one added later, for $75.

One person, one role.

Buy it

Bundles and your own plan

Take every app in a bundle and it costs $15 less on a lifetime key and $5 less on a yearly key. Each app counts toward one bundle. keepcli, when it arrives, counts as three apps.

Study
keeptrack, keepcourses, keeptraining
Everyday
keepconverting, keeptrack, keeplogin
Tools
keepgit, keepmonitoring
Teaching
keepinstruct, keepsetup-instruct, keepcourses

Customize your plan Pick the apps you want and see the price before you buy.

Keys by type

Your own key
One person in one role (student, teacher, editor or viewer). Bought at the prices above, or given to you. Works again after a reinstall.
Class key
One class or team and a set number of people, from 1 to 1000. Each request goes to the server that made the key, and the host approves it. No listed price; ask the owner.
Institution key
Single-use. The owner issues these to a school or company, which hands them to its instructors. Not sold here.

“Buy it” opens the owner's Stripe checkout in a new tab; after payment, keepkeys' Stripe sync issues your key and emails it to you. For a bundle or your own pick of apps, use “Customize your plan”: it shows the total and carries your apps to the checkout, and the amount is checked against the same price table.

Get started

Three steps to your first sign-in.

  1. Get a key

    Buy a key for each teacher, or ask the owner for institution keys to hand to your instructors. Teachers then request class keys for their classes, which cover their students.

    Class key
    Made for one class or team and a set number of people. The host approves each request.
    Your own key
    One person, one role. Bought from the owner (yearly or lifetime; see pricing) or given to you. Works again after a reinstall.
    Institution key
    Single-use, issued by the owner to a school or company for its instructors. Not sold.

  2. Install the apps

    Windows, macOS or Linux. Choose what you want below and the exact command appears; the installer skips whatever is already there.

    No install on a phone or tablet: open the server address shown in keepinstruct's Server panel in your browser (it looks like https://192.168.1.5:8443) and sign in. What the server shares with your role is there; the desktop apps and the local models need an install.

    Already have part of it? tools/install-suite.sh --check lists every step and what is present, without touching the network.

    Fully customizable

    Say what you want in your own words. Which apps, models or not, which class, which machine.

    1. Draft it in the box, then copy it (click in the box, Ctrl+A, Ctrl+C).
    2. On your computer, run tools/install-suite.sh --custom. It asks for your description; paste it there.
    3. Fable, running on your own Claude subscription on your own computer, turns it into an install plan: which apps, which steps, which downloads. You confirm the plan before anything is installed.

    No Claude sign-in? The installer offers a checklist instead. What you type here stays in the box and is sent nowhere.

  3. Sign in

    Use your username and the one-time code your host gave you to choose a password of at least 12 characters. From then on, your username and password are all you need.

    Log in See the apps

Works cited

The studies behind the ADHD notes.

Numbered where they are used above. Each link opens the paper at its publisher.

  1. Barkley, R. A. (1997). Behavioral inhibition, sustained attention, and executive functions: Constructing a unifying theory of ADHD. Psychological Bulletin, 121(1), 65–94. doi:10.1037/0033-2909.121.1.65 ↩
  2. Gollwitzer, P. M., & Sheeran, P. (2006). Implementation intentions and goal achievement: A meta-analysis of effects and processes. Advances in Experimental Social Psychology, 38, 69–119. doi:10.1016/S0065-2601(06)38002-1 ↩
  3. Gawrilow, C., & Gollwitzer, P. M. (2008). Implementation intentions facilitate response inhibition in children with ADHD. Cognitive Therapy and Research, 32(2), 261–280. doi:10.1007/s10608-007-9150-1 ↩
  4. Harkin, B., Webb, T. L., Chang, B. P. I., Prestwich, A., Conner, M., Kellar, I., Benn, Y., & Sheeran, P. (2016). Does monitoring goal progress promote goal attainment? A meta-analysis of the experimental evidence. Psychological Bulletin, 142(2), 198–229. doi:10.1037/bul0000025 ↩
  5. Buehler, R., Griffin, D., & Ross, M. (1994). Exploring the “planning fallacy”: Why people underestimate their task completion times. Journal of Personality and Social Psychology, 67(3), 366–381. doi:10.1037/0022-3514.67.3.366 ↩
  6. Barkley, R. A., Murphy, K. R., & Bush, T. (2001). Time perception and reproduction in young adults with attention deficit hyperactivity disorder. Neuropsychology, 15(3), 351–360. doi:10.1037/0894-4105.15.3.351 ↩
  7. Martinussen, R., Hayden, J., Hogg-Johnson, S., & Tannock, R. (2005). A meta-analysis of working memory impairments in children with attention-deficit/hyperactivity disorder. Journal of the American Academy of Child and Adolescent Psychiatry, 44(4), 377–384. doi:10.1097/01.chi.0000153228.72591.73 ↩