A good idea showed up while you were walking to lunch. You think you opened your phone and typed it into something — Apple Notes, maybe, or your work-Notion, or a draft text to yourself. A week later, you can't find it, and you can't quite remember what it was either. You vaguely remember being excited.
This happens because most people capture ideas across a half-dozen apps that don't talk to each other. Apple Notes for the personal stuff. The work app for the work stuff. Drafts in your messages. Photos of whiteboards in your camera roll. Voice memos in the voice memos app. Bookmarks in your browser. Open tabs that mean "I'll come back to this." Each app is good at one thing. Together they're a leaky bucket.
The fix isn't a sixth app. It's one vault that holds everything — text, voice, PDFs, images, links — and an agent that can find it later regardless of how you captured it. Below is how that actually plays out, and what changes when ideas stop dying in the gaps.
Why the multi-app problem is structural, not behavioral
Most people blame themselves for losing ideas. The honest truth is that the architecture of multi-app capture guarantees loss, no matter how disciplined you are. Three reasons.
First, the classification overhead — every thought makes you decide which app to put it in, before you write it down. Personal? Work? Reading? A reminder? That decision is small per thought and large per day, and it's the friction that kills most capture.
Second, the search asymmetry — you can search inside one app, but not across them. So even when you wrote it down somewhere, finding it requires remembering which app you used. Memory of where you put a thought is roughly as fragile as memory of the thought itself.
Third, the cross-cutting blindness — most useful ideas connect things across categories. The work problem that the weekend reading shed light on. The personal habit that bears on a professional decision. Multi-app setups can't see across the boundary, so the connections are invisible. (This is the case Why Your Notes App Should Be the Same App for Work and Life makes at length — same root cause.)
One vault, capture in whatever shape the moment wants
In Docapybara, the vault holds everything in one searchable surface. Text notes, voice recordings (with speaker labels via the audio recording feature), PDFs (auto-converted to markdown via docstrange so the agent can read them), images, links — all on the same pages or page trees, all searchable by the same agent.
That means the morning thought drops into whatever page is open, or into a daily Inbox page, or as a thirty-second voice memo. The agent files it later if you want; you don't have to file at the moment of capture.
Pages nest indefinitely, OneNote-style, so the vault grows the way your thinking grows. There's no rigid structure to learn. Just open it, drop the thing, close it.
Voice — the channel that fixes most of the leakage
Voice is the channel that the multi-app setup most often loses. People talk through ideas all the time — to themselves, to a co-worker, to a friend on a walk. The good ideas often surface in conversation. They almost never get written down.
The audio recording in-app fixes most of this. Tap record, talk for thirty seconds (or three minutes), the transcript with speaker labels lands on the page. Now the conversation that gave you the idea is searchable, alongside everything else.
For the recurring conversations specifically — the weekly call with a friend who thinks well, the 1:1 with the colleague you co-think with, the conversation with your partner about long-running plans — voice capture turns ephemeral talk into compounding material. (For the broader capture-as-habit shape, The Capture Habit: Remembering the Things That Actually Matter covers it.)
The agent searches what you couldn't search before
The mechanic that makes one-vault actually work is the agent. Without it, one vault becomes one folder full of mixed content with no good way to query. With it, you can ask in plain English.
"What did I capture last week about the [project] question?" — answered, regardless of whether you typed it, said it, or screenshot-and-pasted it.
"Has the topic of [X] come up across any of my notes in the last six months?" — answered. The agent reads across pages and surfaces what's relevant.
"For the post I'm working on, pull anything from my recent reading or voice notes that's relevant." — answered. Drafts you start with the agent's pull are grounded in your own material, not in something it made up.
Capy (the agent) has 27 tools across the vault, including searches across pages, child pages, and uploaded files. The mechanic — agent acts on documents, not just chats about them — is what we built Docapybara around. We wrote about the broader differentiator at Claude Code for Documents if you want the full pitch.
The migration question — moving from six apps to one
The migration question is the one that stops most people from switching. Honest answer: don't migrate everything. The old notes mostly stay where they are; the vault becomes the home for new capture from this week forward.
In practice, most people find that within a month, the vault has more useful, recent material in it than the old apps did. The old apps quietly retire. By month three, you'd only open them to pull a specific thing you remember being there.
For the things you do want to migrate — the active client list, the long-running project notes, the reading list — drop them on relevant pages over the first few weeks. The agent can help: "I'm pasting an old project notes export. Sort it into the right pages in my vault — keep the original chronology."
What changes after a few weeks
Three things tend to shift after a few weeks of one-vault.
The first is that the "where did I put that" question stops showing up. You always put it in the vault. The agent finds it.
The second is that the capture rate goes up. Things you wouldn't have bothered to write down (because the choose-an-app overhead was too much for the thought) now get captured. A year later, the vault holds material you wouldn't otherwise have.
The third is the unexpected one — synthesis becomes possible. The agent can pull across months of capture and surface patterns you couldn't see in real time. "What questions have I kept coming back to in the last quarter?" gets a real answer. Hard to do in any setup that fragments capture across apps.
(For the long-arc archive shape this becomes part of, Turn Casual Captures Into a Searchable Life Archive covers it.)
What about the apps that are good at one thing?
Some of the multi-app stack is worth keeping. Calendars are calendars; the vault doesn't replace them. Email is email; the vault doesn't replace inboxes. Specialized tools you use professionally (Linear, Figma, your accounting tool, your video-editing software) keep doing what they do.
What the vault replaces is the notes layer — the place where you currently put text, voice, links, and quick captures across half a dozen places that don't talk to each other. That layer is the leakiest one in most people's stack, and it's the one that benefits most from consolidation. (For the honest comparison to other notes tools specifically, Looking for an Obsidian alternative? Here's when Docapybara fits — and when it doesn't covers when the trade-off makes sense and when it doesn't.)
A starter shape for week one
If you're moving from "ideas die in the gap between six apps" to a vault, this is what we'd suggest:
- Inbox — one page for raw daily capture. Don't worry about filing.
- Today — one page that holds today's open items. Refresh each morning.
- Reading — one page or page tree for what you're consuming
- Conversations — one page tree, one child per person you talk with regularly
- Projects — one parent page, one child per active thing
That's it. Five pages. The agent does the cross-cutting search across all of them.
For the first two weeks, the rule is "if I'd normally type it into Apple Notes or Drafts or work-Notion, type it here instead." Voice capture for the longer-form stuff. After two weeks, look at what's accumulated. The structure that fits your life will become obvious; you don't have to design it up front.
The point of one vault isn't ideological — it's practical. Less app-switching, less classification overhead, fewer ideas dying in the gap. (For the work-meets-life version specifically, The Same App for Sprint Planning and Grocery Shopping walks through it concretely.)
Try Docapybara free — drop one week of capture into the same vault, and see how the friction shifts.