You log your trades — the system finds what they have in common and drafts a bot that enters the same way.
How it works
Select 3 to 50 journal entries with the same setup and choose a timeframe.
For every entry, indicators are recalculated at the moment of entry on the last closed candle — by the same engine that runs backtests. So the numbers in the rules mean exactly what they mean in the backtest, and older entries work too.
The system looks for what they share: events (candlestick patterns, Price Action, SMC), oscillator ranges, price versus moving averages, trend direction.
Stop and target are the medians of your entries; direction follows where the stop is placed.
You get a filled-in bot builder form with a report and a “Save and backtest” button.
Why not just “whatever matched”
Three entries can also share something that is true almost all the time — say, “RSI between 30 and 60”. So every condition passes a shift significance test: your whole set of entries is shifted through history while keeping the gaps between them, and we count how often the condition would match all entries by chance. Only conditions with a chance match below 1% make it into the rule.
The report shows, for every condition, on what share of bars it occurs and its chance of a random match.
How accurate it is
We tested the feature on 25,597 synthetic journals over real crypto, stock and forex history. On unseen data, the bot's trades match the real setup's trades with F1 0.44 — versus 0.12 for random entries and 0.36 for a simple threshold method; with three entries, 0.37 versus 0.21. Method and limitations are in the open study.
Worth keeping in mind
Fundamentals, market sentiment, on-chain and macro data lack deep enough history — the report honestly lists which analytics factors did not make it into the bot.
With 8 or more entries, the bot repeats the setup noticeably more accurately.
A draft is not a finished strategy: check it with a backtest and optimization.
The screener is always free. The bot builder, journal and analytics are part of the subscription, with a free 3-day trial.
Not individual investment advice. Trading involves the risk of losing money.