Body data in.Life decisions out.

A local AI lab that runs experiments on your life — and tells you, with proof, what actually makes it better.

Scroll
0%
capture runs on-device
0
local AI models · no uploads
n=0
the only sample size that matters
$0
servers · it's your folder
The problem

Every tracker optimised a number. Your life fell out of frame.

Recovery. HRV. Steps. People grind a score up while the thing it was supposed to stand for — a life that feels good to be inside — quietly slips. That's Goodhart's law wearing a fitness band: when a measure becomes the target, it stops measuring anything that matters.

Why it exists · the point

You don't want more data. You want a life that's actually good.

Longevity, recovery, sleep — all of it is a means. The end is whether your days are worth having. This system keeps that end in the frame on purpose, because it's the one thing a tracker structurally can't: a wearable doesn't know what your good life looks like. You do — so you name it, and everything else is held accountable to it.

The top metric isn't a body score.
Is my life actually better?

Body Movement Food Mind Sleep Work Emotion Connection Meaning Autonomy Growth People
recovery ↗ your life — flat PROXY UP, LIFE FLAT = A FAIL

When the body score climbs but life doesn't, that isn't a win to celebrate. It's metric tyranny — and the system flags it instead of congratulating you.

Built for people who've tried enough supplements, trackers, and protocols to know the hard question isn't "what's optimal?" but "what's optimal for me, in the life I actually live?"

The science

Personal science, not correlations.

Correlations are guesses — "you sleep worse when you drink" can't tell if the drink did it or a hard day caused both. So the harness borrows the method clinical medicine uses to decide if something works for one specific person: the n-of-1 trial, pre-registered, controlled, judged against a line drawn before the data arrives.

success criterion · set before the data baseline change one variable MERGE → RULE
01

Hypothesis

A specific causal claim.

02

One variable

Hold everything else. The rule everyone breaks.

03

Pre-set criterion

Written before the data, so you can't fool yourself.

04

Verdict → rule

Merge it, or revert and log why. A washout confirms.

And every claim the agent makes carries its epistemic weight — observation split from interpretation, each line cited to its source:

FACT · measured INFERENCE · reasoned ASSUMPTION · stated UNKNOWN · admitted RISK · flagged

“Creatine moved nothing for you in three weeks — stop paying for it.” — the kind of sentence no wearable will ever say

Grounded, not vibes

Standing in evidence-based medicine.

Not a medical device — and not invented. The discipline it enforces is the one clinical science uses to decide whether something works for one specific person.

OCEBM LEVEL 1 · FOR YOU n-of-1 Systematic reviews Population RCTs Cohort · case · opinion
For a population, RCTs rank higher. For deciding what works for you, the n-of-1 trial is the top tier.
n-of-1 trials

Single-subject design from personalised medicine: pre-registered, one variable, washout, verdict.

Outcome over surrogate

Anchored on a patient-reported outcome (PRO), not a proxy like HRV. Validated tradition (WHO-5).

Evidence grading

Every claim labelled fact / inference / assumption and cited — observation split from interpretation.

Momentary capture

Logging in the moment is ecological momentary assessment — it beats recalled, reconstructed diaries.

The receipts

Full clinician brief, limitations, and the rest of the references — in METHODOLOGY.md ↗

Try it

Run an experiment. Watch it become a rule.

Pick a change. The harness holds one variable, measures against a line set in advance, and returns a verdict — merge, or honestly, revert.

success criterion baseline change one variable
How it works

Capture loose. Reason tight. Decide.

01
Capture

Live, on the fly

Text, voice, or a photo — selfie, food, stool auto-sorted on-device. Several a day. Zero forms.

02
Synthesise

Cited, on-device memory

Hybrid RAG retrieves your real history; the agent finds patterns you can’t see and writes them down with sources.

03
Decide

Experiments become rules

A hypothesis turns into an n-of-1; a verdict turns into a standing rule — knowledge that compounds.

One camera, three diaries

Snap anything. It sorts itself, on-device.

Selfie
skin · fluid · the spark
Food
plate · protein · timing
TYPE 1 · · · 7
Gut
Bristol scale · colour
CLIP vision routes each photo locally — no caption, nothing uploaded, never diagnostic
What it does

A system, not a chatbot.

Capture

Text, voice, photo — sorted on-device

Talk, type, or snap. Selfie, food, and stool photos route themselves into the right diary with on-device vision — no captions, no forms, nothing uploaded.

Rhythm

A gentle daily nudge

Morning brief, a question or two across life domains, an evening recap. Mute it anytime.

Weekly

The verdict

The agent reads your week and scores it against your north-star.

Causal

n-of-1 experiments

Run, extend, or log a lapse. One active at a time — truly one variable, so the effect is attributable.

Baseline

Knows your habits

Tell it what’s already standing — 15k steps, omega for years — and it won’t re-discover them.

Horizons

Week · month · year

Watch quality of life move over seasons as data compounds.

The engineering

Local-first. No servers. No surveillance.

Your whole life is a folder of plain Markdown you own — a layered pyramid where evidence flows up and nothing is asserted without a trail.

06 outputs 05 decisions · n-of-1 04 synthesis 03 wiki 02 sources · fact / inference 01 raw · you + Whoop 00 context · north-star
On-device voice
Whisper (ONNX) transcribes locally — audio never leaves the device.
On-device vision
CLIP zero-shot routes every photo into the right diary, in milliseconds.
Hybrid RAG
multilingual-e5 + sqlite-vec + FTS5, fused with RRF — your history, cited, hosted on your laptop.
Agent surface
A local MCP server exposes search & synthesis to any client — reasoning over the pyramid, not guesses.
Evidence labels
Every claim is fact / inference / assumption with a source — observation split from interpretation.
Quietly automatic
Race-safe OAuth, a morning brief via launchd, a kb-doctor health-check.
See the difference

Whoop tracks your body. This tracks your life.

A wearable is a brilliant sensor — for physiology. But the drivers of how a day actually feels live mostly outside its straps. Flip it:

4 of 12 life domains in frame
Recovery
body
HRV
body
Sleep
body
Strain
body
Mood
emotion
Energy
vitality
Work
focus
People
connection
Food
intake
Meaning
purpose
Autonomy
agency
Growth
flow
Output
A recovery score. The things that most move how you feel — work, people, meaning — stay invisible.
The killer question

Find your real lever.

Toggle what a week held. Watch what actually moves "is life better?" — and what just looks like progress.

Is life better this week?
The honest number

What it costs to run.

Capture is free — everything on-device. The only spend is the agent's analysis. Move the dials:

7
$0
included in your plan — no per-photo charge
Honest answers

The questions you'd actually ask.

Is my data private?+

Capture is on-device: voice→text, photo sorting, and search all run locally, and your files live on your disk (photos and sensitive diaries are never pushed to git). Two things do leave, by design: messages transit Telegram, and deep analysis uses Claude when you ask the agent. Nothing is sold or posted.

What does it cost to run?+

The capture loop is free — bot, on-device models, and storage cost nothing. The only spend is the weekly analysis: included if you run the agent on a Claude subscription, or cents per session on a raw API key. A photo costs well under a cent.

Is it a medical device?+

No. The agent describes and compares; it never diagnoses. Any worrying sign — from a stool photo or anything else — resolves to a single recommendation: see a specialist.

Do I have to journal every day?+

No. Body data arrives on its own; the rest is whatever you feel like sharing. Skip days freely — it's a tool for your life, not another job.

Who is it for?+

People who've tried enough supplements, trackers, and protocols to want what works for them, in the life they actually live. Not for streaks, badges, or a prettier sleep chart.

Own your data. Run the experiment.

Open-source, local-first, built to generalise to any self-quantifier.