An app blocker with a quiz you can’t skip

The loop, in four movements

Opening a wrapped app starts a sequence that takes a few seconds. It is worth describing precisely, because the difference between a blocker that teaches and one that only obstructs is entirely in the third movement.

1. You open the app. Popcue notices.

   Popcue holds Android’s usage access permission, which is how it knows that one of the apps you selected has just come to the foreground. It reads the fact that an app started. It reads nothing about what happens inside it: no messages, no browsing history, no location, no clipboard.

2. A card is already on screen. There is no loading state.

   Cards are generated and buffered ahead of need, so when the interrupt fires the card is already rendered. This is a deliberate constraint on the design rather than an optimisation: a spinner would give the interrupt a moment of ambiguity, and an ambiguous moment is where a skip button would go.

3. You answer it. Two taps, and the second one is the point.

   Tap once to reveal the answer. Tap again to rate your recall as Got it or Still learning. Both taps are required. There is no skip button, no dismissal gesture, and no timeout that lets the app through on its own.

   The second tap is the mechanic that makes this a teaching tool rather than a speed bump. A pause you can breathe through teaches nothing, because nothing was retrieved. Rating the attempt is what turns the interruption into a data point, and the data point is what the scheduler needs in order to decide when to ask you again.

4. The app opens. Your recall is scheduled.

   The interrupt clears and the app you wanted opens. Behind the scenes the rating you just gave has updated the card’s SM-2 state, which sets the next date that card comes back (sooner if you missed it, later if you got it. You spent a few seconds instead of a few seconds, and the difference compounds over weeks.

The ratio is a ratio, not a quota

Most blockers ask for a daily budget: twenty minutes, or one hundred opens, and then the app is simply gone for the rest of the day. Popcue asks for a per-launch ratio instead, defaulting to 5 out of 10, and the choice within that ratio is randomized.

Three reasons, and they are the reason the default is 5 rather than 10:

• A blocker that fires every single time stops being a habit and starts being a punishment, and a punishment you have to win against is one you eventually route around.
• Randomized within the ratio means it never feels mechanical. Every fifth open, on a visible counter, becomes something you brace for and start pre-empting. Random does not.
• Rare enough to stay a real interruption. If it fires often it stops being an event, and an event is what makes the recall land.

The number is per app, and any app can be set to zero, paused indefinitely, or given active hours so it never interrupts you between, say, 22:00 and 07:00. Active hours are available per app and per goal.

Why there is no skip button

Popcue has no skip control on either step of the interrupt, and there is no setting that adds one. This is the single decision that most distinguishes it from every other app blocker, and it is worth being straight about the objection: without a skip, a person in a bad moment can be annoyed by Popcue.

The reasoning is short. A breathing pause is a speed bump: it slows you down and you resume almost immediately, having retained nothing. A forced recall is a lesson: you have to produce something, and producing something is what makes it stick. A skip button would convert every card from a lesson into a speed bump, and Popcue would be a slightly more irritating version of the category it is meant to be different from.

What Popcue does instead of a skip is give you control over the conditions rather than over the interrupt:

• Set any app’s ratio to zero and it never fires at all.
• Pause any app for as long as you want, or turn the whole thing off.
• Set active hours per app and per goal.
• Lower the ratio on the apps that frustrate you and keep it high on the ones that do not.

And for the rating itself: “Still learning” is not a failure state and is never punished. It is the expected answer for a card you have not seen in a while, and it simply schedules that card sooner. There is no streak at risk inside the interrupt.

Eight ways to be asked

One question shape is not enough. A vocabulary item wants a cloze, a definition wants a plain question, a diagram only works as an image, and a fact you hold confidently is better tested by being asked to reject it. Popcue renders eight card types:

What SM-2 does, without the jargon

Every rating you give updates a scheduler for that card. Popcue uses SM-2, a published algorithm from 1987 that is the foundation of Anki and most of its descendants. The mechanics that matter to a user are these:

• A card you rate “Still learning” comes back soon, usually within the same session, and again the next day.
• A card you rate “Got it” gets a longer gap, and the gap grows each time you get it right.
• Each card carries its own difficulty, so a hard card you know well is not scheduled like an easy card you have never failed.

This is what makes the interrupt additive rather than annoying. Without spacing, a card is either never seen again or seen so often it becomes wallpaper. The scheduler means a small number of cards, asked at the right intervals, accumulates over months.

Popcue also shows progress against the scheduler rather than against a raw count. A card is mastered when its interval has grown past a threshold, not when you have pressed a button a certain number of times.

Where the cards come from

This is the part that distinguishes Popcue from the rest of the category, and it is worth being precise about the difference. Every other quiz-gate asks you to get your material into their silo first: you export a CSV, or upload a PDF, or copy text across, and from then on your notes live in two places. Popcue reads from where the material already is.

The connectors are unlimited and free on every tier, including Pro, because serving them costs us nothing. (Todoist is coming soon, so that is three today.) Charging for them would be charging you for your own notes. Read how each connector works, including what leaves your device and what does not.

Setting it up: five steps, about three minutes

1. Name what you’re learning

   A goal, in your own words. A bounded one you can finish, or something ongoing that never quite ends. Everything after this hangs off it.

2. Connect where you already keep notes

   Notion, AnkiDroid, or any MCP-compatible LLM endpoint of your own (Todoist is coming soon). Or skip it and name a topic and let Popcue write the deck. Every card shows where it came from.

3. Pick the apps to wrap

   Any installed app, not a fixed blocklist. The feed, the shorts, the group chat, whatever actually takes you. Each one gets its own ratio and its own hours.

4. Set the ratio, not a lockdown

   Five out of ten launches, randomized, or whatever you want. Set active hours so it never interrupts you at 1am. A pause, not a prison.

5. Grant two permissions

   Usage access, so it knows when one of the apps you picked opens. Draw over other apps, so the card can be the thing on screen. Each is explained before the system dialog, and nothing else is required.

The full setup guide, including the two permissions and how to check them yourself, or join the waitlist to get the build when it ships.

What Popcue does not do

A page about how something works is not much use if it only covers what the thing does. So, plainly:

• It does not block websites. The wrap mechanic is Android-app-shaped today. Browser blocking is on the roadmap. One Sec blocks sites on iOS through a Safari extension, and if that is the thing you need, One Sec is currently the better tool for it.
• There is no iOS version, and none is scheduled. iOS restricts what a third-party app can present over another app, which makes a genuine full-screen interrupt much harder. We would rather ship one platform properly than hint at a date we cannot stand behind.
• It is not clinically validated and makes no health claims. Popcue is built on two well-documented principles from the study literature, that retrieving something beats re-reading it, and that the same effort spread over days beats the same effort in one sitting, but the product is not a medical device, is not ADHD treatment, and is not a substitute for anything a clinician would advise.
• It is not on the Play Store yet. There are no public ratings, reviews or install counts for Popcue, because it has never been downloadable by anyone but us.

Where to go next