Switch to sequential batching to eliminate rate limit hits

1 concurrent + 20s pacing = ~3 batches/min × 2k tokens = 6k TPM,
safely under the 8k output TPM limit. Progress estimate updated to 22s/batch.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

src/PowderCoating.Infrastructure/Services/AiCatalogPriceCheckService.cs

@@ -22,7 +22,7 @@ public class AiCatalogPriceCheckService : IAiCatalogPriceCheckService
 
     private const string Model = "claude-haiku-4-5-20251001";
     private const int BatchSize = 25;
-    private const int MaxConcurrentBatches = 2;
+    private const int MaxConcurrentBatches = 1; // sequential avoids bursting past the 8k output TPM limit
     private const int RateLimitRetrySeconds = 65;
     private const int MinBatchIntervalSeconds = 20; // proactive pacing: ~3 batches/min × ~2k tokens = ~6k TPM, under the 8k limit
 

src/PowderCoating.Web/wwwroot/js/catalog-price-check.js

@@ -11,11 +11,10 @@
     if (!form || !btn || !overlay) return;
 
     // Estimate total seconds based on item count.
-    // Haiku: 2 concurrent batches, ~25s per wave (includes occasional rate-limit waits).
+    // Haiku sequential: 1 batch at a time, ~22s each (API time + 20s pacing gap).
     function estimateDuration(itemCount) {
         var batches = Math.max(1, Math.ceil(itemCount / 25));
-        var waves = Math.ceil(batches / 2);
-        return Math.max(30, waves * 25);
+        return Math.max(30, batches * 22);
     }
 
     // Messages keyed to approximate progress milestones (0–100).