perf(metrics): Reduce token counting batch size for better worker utilization

src/core/metrics/calculateSelectiveFileMetrics.ts

@@ -9,8 +9,12 @@ import type { FileMetrics } from './workers/types.js';
 // Batch size for grouping files into worker tasks to reduce IPC overhead.
 // Each batch is sent as a single message to a worker thread, avoiding
 // per-file round-trip costs (~0.5ms each) that dominate when processing many files.
-// For 991 files: 991 round-trips → 20 batches, saving ~485ms of IPC overhead.
-const METRICS_BATCH_SIZE = 50;
+// A size of 10 keeps individual worker tasks small so that workers become available sooner,
+// enabling overlap between file metrics and output generation.
+// When tokenCountTree is disabled, metrics only processes a small number of top files
+// (e.g., topFilesLength * 10 = 50 by default), so a smaller batch size avoids
+// a single batch monopolizing one worker.
+const METRICS_BATCH_SIZE = 10;
 
 export const calculateSelectiveFileMetrics = async (
   processedFiles: ProcessedFile[],

src/core/metrics/workers/calculateMetricsWorker.ts

@@ -7,7 +7,7 @@ import { freeTokenCounters, getTokenCounter } from '../tokenCounterFactory.js';
  *
  * Supports both single-content and batch modes. Batch mode reduces IPC overhead
  * by processing multiple files per worker round-trip (~0.5ms overhead per round-trip).
- * For 991 files, batching with size 50 reduces round-trips from 991 to 20.
+ * For 991 files, batching with size 10 reduces round-trips from 991 to ~100.
  */
 
 // Initialize logger configuration from workerData at module load time

src/core/security/securityCheck.ts

@@ -21,8 +21,10 @@ export interface SuspiciousFileResult {
 // Batch size for grouping files into worker tasks to reduce IPC overhead.
 // Each batch is sent as a single message to a worker thread, avoiding
 // per-file round-trip costs that dominate when processing many files.
-// A moderate batch size (50) reduces IPC round-trips by ~98% (990 → 20 for a typical repo)
-// while keeping enough batches to utilize all available CPU cores.
+// Security check always processes all files (~1000 in a typical repo), so a batch size of 50
+// already produces ~20 batches — enough to distribute well across available CPU cores.
+// (Unlike metrics, which may process only a small number of top files when tokenCountTree
+// is disabled, and needs a smaller batch size to avoid one batch monopolizing a worker.)
 const BATCH_SIZE = 50;
 
 export const runSecurityCheck = async (