Most backtests fill you at prices you’d never get. We don’t. Describe a strategy, or let your AI (like Claude) do it, and see if it holds up.
Pick one everyone talks about and watch our engine grade it on real prices, real spreads and history the rules never saw. Nothing is withheld before the result.
A strong stock falls hard for a few days. You buy the bounce.
The rule we’ll testDown 10% or more in 3 days, but still above its 200-day average. Buy the shares, hold 10 days.
How it works
Paste your prompt into any AI — Claude, Gemini, Codex, GPT. Terminal or desktop. Connected. From then on it writes the strategy, tests it, and runs the ones you say yes to.
Buy at the ask, sell at the bid, never the middle, because nobody fills you there. Every fill is written down at the price it really paid.
Our engine’s read on your strategy, tested on years it never saw and printed beside your own choice. It is an opinion, never a gate. You decide what runs.
Your bots run the whole session, exactly as you designed them — open to close, without you watching.
A stop rule sells at the bid, on the bar where it fired — never on the friendlier one an hour later. Protection you can check afterwards.