For prop-firm and personal-account traders

Was it the model, or was it you?

Every setup reaching you has already argued for itself — multi-agent debate, replay-validated against thousands of past bars. Then Sentio does the part no journal can: it keeps scoring that setup whether you take it or not.

So when a week goes badly there are two curves, not one — what the rules did, and what you did. The distance between them is the only number that tells you which one to fix: the model, or the "one more" at 15:40.

The four honest failure modes

The trades you don't journal about.

These are the ones that actually kill accounts. Not the C+ setup you skipped. The four A- setups you never should have taken.

✕ Losses 3–5 of the day

You already broke your own rule at loss #3. Now you're "in size to make it back."

✓

Daily-cap gate fires at trade 3. The scanner stops surfacing new setups until tomorrow. No willpower required — the button just isn't there.

✕ The revenge trade at 14:47

Stopped out at 14:41. Sitting there staring at the chart. Sixteen minutes later you're double-size in the opposite direction.

✓

Post-loss cooldown blocks new alerts + the decision dialog for a window you set. The next signal is muted until the timer clears.

✕ Size creep on a green day

Two wins in. Suddenly the 2-lot feels small. You go 4. That's the one that gives it all back.

✓

Sizing engine locks to your MAX RISK cap per trade. It doesn't care how good you felt about the last two.

✕ Trading the wrong session

Your win rate on 2pm–4pm is terrible and you know it. But you're bored and there's an "obvious" setup.

✓

Per-model time-window mutes. Mute Unicorn between 14:00–16:00 ET. Setups don't surface. Boredom stays boredom.

No broker integration

Sentio never touches your orders.

There's no API key to hand over, no positions being read, no fills coming in. Sentio surfaces the setup + reasoning on one screen — you place the trade on your broker's platform on the other. Prop firm, funded challenge, personal account, doesn't matter which front-end you use.

Sentio · this screen Structure + guardrails Every setup with the reasoning trail. Discipline rules enforced. Replay of every trigger.
Your broker · other screen Execution You place the order. You take the fill. You manage the position. Sentio is not in the loop.

This separation isn't a limitation — it's the whole architecture. It's what lets Sentio operate as a publisher (CFTC Rule 4.14(a)(9)) instead of a Commodity Trading Advisor. You keep the decision. We keep the analysis. Neither of us hands the other the wheel.

The mechanism

The rules run themselves.

Every trader has a version of themselves at 9am who writes good rules — three trades max, no revenge, halve size after two losses — and a version at 2:47pm who overrides them. Sentio is written for the 9am version. Your discipline layer is plain English on the way in and enforced code on the way out. The scanner doesn't surface a setup that fails a gate you wrote. The alert never fires. The trade never appears.

You still do the work — you find the setup, you size it, you execute on your broker. Sentio's job is to keep 2:47pm you from having a button to press. Not popups. Not friction prompts. The button just isn't there.