The number measures
the process.
The system protects it
while you trade.
Two halves of one engine. How the Kairos Score is built — and how the operating states stop you before damage, not after.
Not a log that tells you how it went. A system that steps in while you decide.
How the number
is born.
A journal records what happened. The Kairos Score measures how you decided.
We measure the idea.
Not the single trade.
Almost everyone judges the single trade. We judge the idea. A thesis can live in a single entry or in five tranches: the system aggregates them via VWAP and treats them as one.
One idea, one grade — because it’s the idea, not the single leg, the place where you decided.
Decision, execution,
risk.
Decision — the quality of what you wrote before trading: thesis, invalidation, plan.
Execution — how much the facts respected the plan: timing, exits, partials.
Risk — stop loss present and never widened, size within limits, daily stop respected.
They don’t weigh the same, and the weight is part of the method: decision and risk count more than execution.
Fifteen criteria
behind three numbers.
Behind the three dimensions, fifteen distinct criteria are at work. Most are computed by the code on objective facts — times, prices, stops, lots — so they aren’t open to opinion. Part is judged by a dedicated AI, at low temperature, on a fixed rubric: the same idea judged twice gives the same grade.
Which the fifteen criteria are and how they weigh, we don’t publish. That’s the rubric — and that’s what makes the score hard to game.
Declared
against done.
You declare the idea before trading. The account, connected read-only, records what you actually did. An AI compares the two tracks and looks for deviations.
But it only penalises the ones you didn’t declare. Changing the plan and writing it is discipline; changing it secretly is the problem.
Honesty doesn’t mean perfection: it means what you tell matches what you do.
P&L doesn’t enter.
Ever.
In none of the fifteen criteria is there profit. An idea that loses with a clean process is worth more than one that wins by chance.
It’s the rule that separates a discipline system from a leaderboard of luck.
Every idea
leaves a grade.
The score appears after the first five recorded trades. From there on, every closed idea leaves a grade.
And the grades, summed, become your route.
We measure the idea.
Not the single trade.
Almost everyone judges the single trade. We judge the idea. A thesis can live in a single entry or in five tranches: the system aggregates them via VWAP and treats them as one.
One idea, one grade — because it’s the idea, not the single leg, the place where you decided.
Decision, execution,
risk.
Decision — the quality of what you wrote before trading: thesis, invalidation, plan.
Execution — how much the facts respected the plan: timing, exits, partials.
Risk — stop loss present and never widened, size within limits, daily stop respected.
They don’t weigh the same, and the weight is part of the method: decision and risk count more than execution.
Fifteen criteria
behind three numbers.
Behind the three dimensions, fifteen distinct criteria are at work. Most are computed by the code on objective facts — times, prices, stops, lots — so they aren’t open to opinion. Part is judged by a dedicated AI, at low temperature, on a fixed rubric: the same idea judged twice gives the same grade.
Which the fifteen criteria are and how they weigh, we don’t publish. That’s the rubric — and that’s what makes the score hard to game.
Declared
against done.
You declare the idea before trading. The account, connected read-only, records what you actually did. An AI compares the two tracks and looks for deviations.
But it only penalises the ones you didn’t declare. Changing the plan and writing it is discipline; changing it secretly is the problem.
Honesty doesn’t mean perfection: it means what you tell matches what you do.
P&L doesn’t enter.
Ever.
In none of the fifteen criteria is there profit. An idea that loses with a clean process is worth more than one that wins by chance.
It’s the rule that separates a discipline system from a leaderboard of luck.
Every idea
leaves a grade.
The score appears after the first five recorded trades. From there on, every closed idea leaves a grade.
And the grades, summed, become your route.
The system
doesn’t wait until tonight.
The Kairos Score tells you how it went. The operating states serve to change how it goes — and they don’t live at end of day: they’re active while you trade. This is where the same number flips, and from verdict it becomes a brake.
How the states
fire.
The number tells you how it went. The states are there to change how it goes. And they don’t wait until evening.
Three gears.
The system shifts them.
Normal — full operations. The day’s risk is intact, controls are active. You’re respecting the method.
Throttle — reduced operations. Max size halves, ideas are counted, only structural setups remain. It doesn’t stop you: it shifts you down a gear.
Break-Glass — operations suspended. You crossed a red line. No new ideas until you complete the re-entry protocol.
The system shifts them, based on your process — not you. Try it: touch the gears.
Objective tripwires,
not feelings.
The states don’t fire on intuition. Behind them is a series of objective tripwires, checked by the code on the day’s facts: a missing stop loss, a stop widened after entry, too many ideas in a day, the daily stop breached, having traded on a declared pause day.
They are deviations of equal severity — the system doesn’t issue automatic life sentences. They accumulate, and when the day picture worsens past the threshold, the gear shifts.
The point isn’t the number: it’s that the limit fires by itself, in real time, without you having to remember it.
Live,
not post-mortem.
A journal gives you the report at end of day, when damage is done and all that’s left is to write it up. Kairos works earlier: it reads the day as it happens and, when the process degrades, tightens the constraints on the next trade — not on next month.
No sermons, no motivational lines. An operating constraint applied at the exact moment it’s needed.
A brake that
brings you back on track.
When Break-Glass fires, the system doesn’t just lock you out. It asks you to go through a short protocol: close the positions, complete the day close, and answer three questions — what the cause is, what the countermeasure is, by which criterion you re-enter.
Then a cooldown time, calibrated on whether it’s the first time or the umpteenth. Once the protocol is complete you don’t go straight back to full operations: you re-enter in Throttle and climb from there, with clean days.
Recording is not punishing.
Three gears.
The system shifts them.
Normal — full operations. The day’s risk is intact, controls are active. You’re respecting the method.
Throttle — reduced operations. Max size halves, ideas are counted, only structural setups remain. It doesn’t stop you: it shifts you down a gear.
Break-Glass — operations suspended. You crossed a red line. No new ideas until you complete the re-entry protocol.
The system shifts them, based on your process — not you. Try it: touch the gears.
Objective tripwires,
not feelings.
The states don’t fire on intuition. Behind them is a series of objective tripwires, checked by the code on the day’s facts: a missing stop loss, a stop widened after entry, too many ideas in a day, the daily stop breached, having traded on a declared pause day.
They are deviations of equal severity — the system doesn’t issue automatic life sentences. They accumulate, and when the day picture worsens past the threshold, the gear shifts.
The point isn’t the number: it’s that the limit fires by itself, in real time, without you having to remember it.
Live,
not post-mortem.
A journal gives you the report at end of day, when damage is done and all that’s left is to write it up. Kairos works earlier: it reads the day as it happens and, when the process degrades, tightens the constraints on the next trade — not on next month.
No sermons, no motivational lines. An operating constraint applied at the exact moment it’s needed.
A brake that
brings you back on track.
When Break-Glass fires, the system doesn’t just lock you out. It asks you to go through a short protocol: close the positions, complete the day close, and answer three questions — what the cause is, what the countermeasure is, by which criterion you re-enter.
Then a cooldown time, calibrated on whether it’s the first time or the umpteenth. Once the protocol is complete you don’t go straight back to full operations: you re-enter in Throttle and climb from there, with clean days.
Recording is not punishing.
The number measures.
The states protect.
Together they’re the engine that turns trading from outcome into process — and the process into something you can see, every day, as a number.