What do you want to talk about?
Ten conversations — ten models. Each one enters a different room of your life and asks what runs you there. The questions adapt to your answers — no two people get the same conversation — and every one scores on the same eight-pattern engine, so the lab can compare a Work default against a Love default and learn what a single life is made of.
Two modes, one instrument. The worst version is the default you fall to under pressure. The best version is the set of conditions your peak runs on — and what to do about them.
The Mirror
The whole picture. Your master pattern.
Work
What runs you between nine and five.
Money
What your hands do when no one watches the account.
Love
Who shows up when it isn't going well.
The 2AM Conversation
Who you are when no one will ever know.
Family
The original pattern, where you learned it.
Your Circle
Who you become around your people.
The Body
The first place the pattern shows.
Ambition
What you actually do about the big want.
The Past
The version of you that still votes.
How it works
You answer by tapping, and sometimes by typing — the typed answers matter; the lab studies language as well as choices. After each answer, the conversation reflects what you said before it asks anything else. It is not collecting a form. It is listening for the gap between the person you present and the one who acts.
When it has enough, it stops. “This conversation has ended.” Your case file opens: your pattern, your depth score, and a day reconstructed inside your default. There is no open chat afterwards — other conversations are the way back in. That boundary is deliberate; an instrument that never ends measures nothing.
Why your answers matter
Every conversation can also speak in three registers — Straight (plain words, real-life moments), Standard (as written), and Deep (the conceptual register). The register changes the language, never the instrument: the same questions underneath, the same scoring, and the register is stamped on your case file next to the model version. The Mirror speaks all three today; the rest are being calibrated register by register, in the open.
Each model carries a version. As the lab learns, the models improve — sharper questions, deeper branching, better-calibrated depth — and every result is stamped with the model that produced it, so the science stays comparable across generations.
Every closed conversation files one anonymous result in the Pattern Index — pattern, depth, conversation, country. Nothing with your name on it, ever. Your file is the value you take; the aggregate is the science the lab publishes openly. People come for the verdict on themselves. The dataset is what their honesty leaves behind.
Your case file
The report a conversation opens: the pattern, the depth score, and a day reconstructed inside your default.
The depth score
How deep the pattern runs, on a 5–99 scale — and why two people with the same pattern live very different lives.
Your best file
The document Best Mode opens: your peak profile, read from your best days — and The Prescription, the part you act on.