01
Intent
What the strategy meant to do.
Post-trade audit
Before you scale the bot, know whether live edge survived execution — or only the backtest.
Evidence between strategy and capital. Intent, gates, orders, fills — then edge, execution, or luck.
Chain
Fill the backtest → live gap.
Assume a system fired. Reconcile intent, gates, orders, and fills — then decide if live still matches the plan.
02
Gate
Rule blocks. Missed price ≠ failed execution.
03
Order
Lifecycle only when the log proves it.
04
Fill
SnapTrade / broker clocks. No fake midnight.
05
Verdict
Claim, effect, limit, next test.
Pillars
What matters before you add size.
Capture is logistics. The product is the audit.
Sim-to-real
Did live match the backtest?
Daily drift and missed entries — before you scale a simulated curve.
Execution
Gates and fills, separated.
Opportunity → block → market order → fill. One denominator each.
Honesty
No vanity stats.
Tests with fixed seeds. Small samples stay insufficient.
VERDICT · FRAGILE
Claim: live expectancy rides few winners
Evidence: effect · sample · CI · trades
Limit: regime mix may confound
Next: freeze rule · 20 holdout
Compare
Before you scale: the checklist.
What a bot trader must know that a diary never answers.
Pricing
Free core. Paid continuous audit.
Local trust is free. Pro sits beside the automation you already pay for.
FAQ