A small fix, but the honest kind. Your backtest already runs on a second-by-second replay of the real market — now the exit time you see matches the exit that actually happened.
When a trade closed at, say, 09:22, the engine knew the exact moment — but the receipt used to show it snapped to the candle boundary (09:30 or 09:00 on a 30-minute chart). The numbers were always right; only the displayed time was rounded. If you looked closely, you might reasonably wonder why a second-by-second engine spoke in half-hours.
The exit position the engine had computed all along now rides along to the receipt. Hover a close marker on the chart, or open the CSV, and you'll see the real exit time, to the second. Nothing about the results changed — the same trades, the same prices — the clock just stopped rounding.
Older cards keep showing the candle time until they're recomputed; new ones carry the precise time from the start.