Import, two minutes
Drop in the CSV your exchange exports, or connect a read-only API key once. Then forget about it.
Import your trade history and it reviews every futures trade for you: where the money goes, which habit costs the most, what to do next. You don't run the review — the conclusions come to you.
An email address is all it takes to start — no card. Built around Binance USDT-margined futures.
You can recite your win rate. You can't name your most expensive habit.
This is the behavior diagnosis. It groups your trades by habit, works out how much worse each group did than the rest of your trading, and ranks them by that gap.

最贵的错误是「亏损中加仓」—— 比不带这个毛病时合计多亏 24,931.64 USDT。
建议 先给这一笔定一个能接受的最大亏损,再决定要不要加。
On a narrow screen only the single most expensive habit is shown; the full table and evidence panel are on desktop.
You import once. It takes everything from there.
Drop in the CSV your exchange exports, or connect a read-only API key once. Then forget about it.
Market replay, P&L breakdown, habit detection — all of it runs on its own. It does the reviewing, not you.
Where the money goes, which habit costs the most, what to do next — delivered to you.
All four aim at the same thing: one fewer repeat of the same mistake.
Everyone is clear-headed looking at the chart afterwards. Getting back to the moment you clicked is the hard part.
Press play and the candles replay one by one — everything past the cursor is kept from you. On one side, the market as it was: candles, order-book depth, long/short sentiment, frame by frame. On the other, you as you were: entries and exits, leverage, open P&L, all on the same chart. What you practise is the judgement you should have had at the time.

You don't control whether the market is kind. You do control whether you keep your own rules.
The behavior index measures discipline, not P&L: whether mistakes repeat, whether the plan was followed, whether the reviews got done — combined into one score that rolls forward with your trading. It doesn't care how much you lost, only whether you held your own line.

You can hold maybe ten symbols in your head. The market has more than eight hundred.
Once you're done reviewing yourself, sweep the whole market — same market data, pointed at the present instead of the past. Hand your conditions to the machine: price change, turnover, open interest, long/short ratio, funding rate, indicators and patterns — up to eight at once.
A cell that can't be judged says so. Data unavailable is not the same as condition not met, and it is never treated as zero.

The few minutes when you're tilted are usually the expensive ones.
When it spots you repeating a pattern that has cost you before, it hands you your own history. It won't tell you to buy, it won't tell you to sell, and it offers no trading advice at all — whatever stops you comes from the mirror.
The in-trade nudge isn't live yet, so this slot deliberately has no interface shot.
14 days of everything, free, no card required. Walk away if it isn't for you — and ask us to delete your data any time.
Start your 14-day free trialEvery conclusion comes with an evidence panel you can inspect. You're free to disagree with it — what you can't do is wonder what it was based on.
Each rule splits your trades into the ones that hit it and the ones that didn't, then shows both: means, win rates, the median gap. “This is bad” on its own is not a conclusion.
“Impact” is not the sum of the losses on the matching trades. It is (mean of hits − mean of non-hits) × number of hits — the answer to “what did this habit itself cost me”.
Rank-test p, win-rate-test p, multiple-comparison correction, and a “at worst, at least this much worse” floor — all four on the panel.
A rule without enough sample stays in “still watching” and never gets promoted to a conclusion. When no difference can be shown, it says so.
A cell with missing data prints “—”. Can't be computed is not zero — that rule is absolute across the product.
At the bottom of the panel: “see the N matching trades →”. That takes you to the actual trades, not a summary.
These are the product's own numbers — symbols covered, candles stored, sampling rate, rules written; every one of them is checkable in the database. We have no user counts or track records to show, so we show none.
The quality of the review itself is identical on both tiers; nothing about it is held back. The difference is the market side: the whole market, scans, and second-level alerts.
The full trade review — see yourself clearly.
$169 billed yearly — save 25%
Start 14-day free trialThe review, plus an entire market workbench.
$349 billed yearly — save 25%
Start 14-day free trial14-day full trial, no card required; when the trial ends or a subscription lapses you drop to read-only (your data stays visible); full no-questions refund within 14 days of a first order. Billed in USD.
No. The CSV statement your exchange exports is enough to start. If you want it to keep up with your later trades on its own, you can add a read-only API key — you do not need, and should never enable, trading or withdrawal permissions. We never request either of them.
No. Not a word. It is a review system, not a signal service: it helps you see your own trading clearly and is not investment advice. See the risk disclosure.
Binance USDT-margined futures for now — one market, done properly.
Two subscription tiers, monthly or yearly, billed in USD — the prices and what each tier includes are in Pricing above. You start with a 14-day full-feature free trial; no card is required and nothing is charged during it.
Read-only access, encrypted in transit; your data is used for your own review only, never sold, never shared. To delete your account and all of its data, contact us through “Submit feedback” in the product or by email; we complete it within 7 business days. See the privacy policy.
Import once, and it shows you where the money goes, which habit costs the most, and what to do next.
Start your 14-day free trial14 days of everything, no card required. By signing up you confirm you have read and agree to the terms of service and privacy policy.