NeuroCognica
Mirrorborn
A reflection engine that treats your own answers as evidence, shows its working, and keeps your history so you can see how you change — rather than handing you a four-letter label and calling it your identity.
Mirrorborn is not a clinical or diagnostic tool. It does not assess, diagnose or treat any medical or psychological condition, and nothing it produces is advice. It is a structured mirror for reflection.
Status: in development. Accounts and the live assessment are not open yet. This page describes what is being built, and it will say so plainly until the day it is real.
How it works
You answer, at length
Thirty open questions with follow-up probes — written answers, not a slider between two adjectives. The instrument is balanced across five categories.
The model reads, it does not decide
A language model tags one question-and-answer pair at a time against a fixed fifty-node taxonomy. It is a constrained reader, deliberately not the authority on who you are.
Deterministic code decides
Every tag is validated against the canonical taxonomy, and station assignment is done by rules you could audit — not by a model's impression of you.
The five stations
| Station | What it holds |
|---|---|
| Dominant | How you actually operate, most of the time |
| Auxiliary | What supports and balances that |
| Shadow | What is disowned or under pressure — the hardest to elicit honestly |
| Aspiring | What you are reaching toward |
| Stress Dynamic | What takes over when you are under load |
Partial truth, shown as partial truth
If the evidence does not support a station, Mirrorborn leaves it empty and says so, with a completeness figure. It will not invent a Shadow because a blank looks unfinished.
That is a real behaviour and not a hypothetical: in current testing the Shadow station frequently comes back empty because the model under-returns those tags. That is a known calibration gap, it is being worked on, and until it is fixed it will show as insufficient evidence rather than as a confident answer.
Your history is the point
Most instruments of this kind give you one result and treat it as a fixed fact about you. Mirrorborn keeps every assessment you take, so the interesting object is the sequence: which stations moved, what strengthened, where completeness improved as you answered more fully.
Your answers are saved as you give them, so you can stop halfway and pick up later on any device.
Your data
- An account is required, so your history has somewhere to live. Sign-in will be through Google.
- Unlike ChronoSophia, this is a hosted service — your answers are processed and stored on servers, not only on your machine. That is a real difference from our desktop product and we will not blur it.
- You can export everything, delete a single assessment, or delete your account and all of its data.
- The full terms will be published before accounts open.