Get in touch→
← Back to lab
ProductReleasedProject 14 / 14

Inner Atlas

A private canvas for mapping how your own mind works: the triggers, beliefs, emotions and behaviours you've uncovered through coaching, and the loops they quietly form.

At a glance

  • The empty screen is the hardest one
  • A bias is both a step and a lens
  • Private by default, permanently

Open Inner Atlas

The question

Insight from coaching or therapy fades fast. A breakthrough becomes a half-remembered phrase in a notebook, and linear notes lose the very thing that made it useful: how one thought leads to the next. Could a map hold that structure, and show you when a chain closes back on itself?

What I built

A calm, dark canvas with six kinds of shape: trigger, thought or belief, emotion, behaviour, bias and value. You connect them, and the app finds every loop on its own, so you can name it and mark the step where you could get out.

The way in is Map a moment: four plain questions about one recent moment, and the chain draws itself as you answer. A pattern library of 47 documented patterns, each with its source, a timeline that replays how the map grew, and a 3D constellation view sit around that.

Inside

How it looks in use.

Role
Sole designer, developer and owner
Built with
React, three.js with custom shaders, IndexedDB
Runs
In the browser, local-first, no account
Status
Released, early
Map a moment: four answers, and the loop closes on its own.
From the flat map to the constellation and back: values float, core beliefs sink.
A whole map: one belief at the centre, two loops running through it.
A whole map: one belief at the centre, two loops running through it.
Open a shape to write in it, mark it as a core belief, weigh the evidence.
Open a shape to write in it, mark it as a core belief, weigh the evidence.
Every loop found automatically, named, with a way out marked.
Every loop found automatically, named, with a way out marked.
The pattern library: documented patterns, each with its source.
The pattern library: documented patterns, each with its source.

Up close

The details that carry it.

A loop, drawn
A loop, drawnThe orange line is the loop closing: what you did made the thought feel more true.
A bias as a lens
A bias as a lensMind-reading sits on the line between a vague message and the belief it sets off.
Evidence on both sides
Evidence on both sidesA core belief keeps a small ledger of what supports it and what doesn't.

Decisions that mattered

  • The empty screen is the hardest oneA blank canvas is intimidating when the subject is yourself, so the first thing you see is a question, not a tool.
  • A bias is both a step and a lensMind-reading can sit in a chain like any other node, or be dropped onto a connection, where it visibly bends what passes through it.
  • Private by default, permanentlyEverything lives in the browser's own storage. No account, no network calls, and export to JSON, an image or a written summary when you want it.

What I learned

The useful question was the last one: did doing that make the thought feel more true? Asking it in plain words turns a list of reactions into a loop you can actually see, and that one moment does more than any feature around it.

Project notes

This project is released, so you can try it right now. It keeps changing as I use it and learn from the people who do.

Something similar in mind?

Most of these started as a question I couldn't let go of. If you have one too, let's talk.

Get in touch→ Back to lab→
To MaxAn ideaWhat would you like to build?I've been wondering whether…I read every message myselfWrite →