We publish what we build next, before we decide
This app is not aiming to be a faster notepad. It is aiming at something else: you organise nothing, and the thought you had is already in the right place. The five things below are the route there. Tell us which one you want first.
What it does today
Capture from iPhone, Apple Watch, Siri (including through AirPods) and widgets — mostly without looking at a screen. What you capture arrives by email and is appended to your Obsidian vault.
Honestly: right now there is essentially one destination. Where a thought ends up barely branches. The five steps below are how that opens up.
The next five
The order is deliberate. We are not starting with "it files things for you". First it only suggests. Then it learns from the corrections you make. And before it learns anything, the rules you write yourself take priority.
Delivery you can see — no more "I thought I sent that"
Proof it was delivered, the reason when it wasn't, the state of a retry, and a searchable log of everything you have sent. You should be able to look and know it arrived.
This was not in the original plan. It is here because counting reviews across 18 Japanese App Store apps put "it vanished" and "it didn't sync" above every single feature request. Whether it really should come first is one of the things the survey below asks.
Suggest a destination — without saving it there
When you finish writing, it offers "this looks like it belongs here". You decide. Nothing moves on its own.
Suggest onlyLearn from your corrections
Every time you override a suggestion, that becomes the record it learns from. Without this, the suggestions never get better.
LearningWrite your own rules
"If it contains this word, send it here." An explicit rule you wrote beats a guess — and behaviour you can predict is faster than behaviour you have to check.
RulesMore ways in — share sheet, Action Button, Mac
An article in Safari, a photo, something shared from another app. And the same place to drop things while you are working on a Mac. Not a separate app per entry point.
CaptureMore ways out — Reminders, Calendar, email and more
"Do this later" goes to Reminders, "next Tuesday" goes to Calendar. You still write the same way; the destination is what branches.
DestinationsWhat we are not building
Worth stating what is missing from the roadmap on purpose. These are not backlog items we never got to — they are decisions not to build.
- A to-do manager. Reminders and dedicated to-do apps already do that better.
- A calendar UI. Your events are easier to read in the calendar you already use.
- A knowledge base or vault browser. That is Obsidian's job, or Notion's.
- A read-later list. We are not adding another screen to read things in.
To be clear: it is not that nobody wants these. In the survey described below, PDF handling, handwriting and folder management all ranked high. We are choosing not to build them knowing that.
This app means to sit between the moment a thought appears and the place it finally lives. Owning the place it lives would put us in competition with it.
The survey
Every question is optional — answer one and send, that still reaches us. No account, no login. The options come from reading memo-app reviews across the Japanese App Store, so if none of them fit, the free-text box is the important one.
Where these five came from
Two places. One is the direction this app has been heading since the start. The other is 7,279 reviews pulled from 18 leading apps on the Japanese App Store in August 2026, with the analysis centred on the 3,541 from the last 24 months.
The clearest result: ratings do not drop over missing features. Speed and stability drew 466 mentions — 259 of them at one or two stars, and it showed up in all 18 apps. Losing or failing to restore data drew 340, with 195 at one or two stars. Both outrank every feature request. People leave because something vanished, didn't sync, or got slow. That is why delivery you can see sits ahead of the five above.
One result ran against expectation. AI came up 105 times: 42 positive against 63 negative or frustrated. "Let me turn the AI off" and "I'd rather have it fast" were the louder half. That is why this page does not talk about AI.
Caveats worth stating: the source is Apple's public review feed, capped at 500 reviews per app, so it is not the full history. And the combined rating count (~993,000) stands in for popularity — it is not a download ranking.