⚡ Bolt: Optimize clipBars to use O(log n) binary search

.jules/bolt.md

@@ -1,3 +1,6 @@
 ## 2024-05-14 - Time-Series O(log n) Search Pattern
 **Learning:** `src/agent/backtestRunner.ts` makes heavy use of `.filter((b) => b.time <= asOf)` to get the latest close price inside loop bounds, changing potentially O(1) loop lookups or O(log n) lookups into repetitive O(n) filtering over large arrays.
 **Action:** Expose `findLastBarIndex` in `src/data/sources/dnsePublic.ts` and use binary search or track indices explicitly in tight loops.
+## 2024-05-14 - Time-Series O(log n) Search Pattern
+**Learning:** `src/data/sources/dnsePublic.ts` previously used a `.filter((b) => b.time <= asOf)` to clip historical arrays. This turned what should be a fast slice operation into an O(n) scan over entire time-series data.
+**Action:** Used the already-available `findLastBarIndex` (which implements binary search) and `.slice()` to truncate the arrays in `O(log n)` time instead, massively speeding up lookups for arrays representing long timelines.

src/data/sources/dnsePublic.ts

@@ -88,7 +88,10 @@ function clipBars(bars: Bar[]): Bar[] {
     asOfClock.getStore()?.asOfSec != null || isAsOfOverridden();
   if (!hasOverride) return bars;
   const asOf = nowSec();
-  return bars.filter((b) => b.time <= asOf);
+
+  // ⚡ Bolt: Optimize O(n) filter to O(log n) binary search for large time-series arrays
+  const idx = findLastBarIndex(bars, asOf);
+  return idx === -1 ? [] : bars.slice(0, idx + 1);
 }
 
 export async function getStockOhlcv(