Skip proxies for Dutchie - datacenter IPs are blocked

Dutchie blocks all our datacenter proxy IPs, returning empty/different
content. Direct connection from pod IP works fine (100 products found).
Added PROXY_SKIP_DOMAINS list for sites that block datacenter IPs.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

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

backend/src/scraper-v2/middlewares.ts

@@ -69,6 +69,20 @@ export class UserAgentMiddleware implements Middleware {
   }
 }
 
+// Domains that should skip proxy (datacenter IPs are blocked)
+const PROXY_SKIP_DOMAINS = [
+  'dutchie.com',
+];
+
+function shouldSkipProxy(url: string): boolean {
+  try {
+    const urlObj = new URL(url);
+    return PROXY_SKIP_DOMAINS.some(domain => urlObj.hostname.includes(domain));
+  } catch {
+    return false;
+  }
+}
+
 /**
  * Proxy Rotation Middleware - uses the central proxy service with timeout handling
  */
@@ -79,6 +93,12 @@ export class ProxyMiddleware implements Middleware {
   private currentProxyId: number | null = null;
 
   async processRequest(request: ScraperRequest): Promise<ScraperRequest> {
+    // Skip proxy for domains that block datacenter IPs
+    if (shouldSkipProxy(request.url)) {
+      logger.info('scraper', `⏭️ Skipping proxy for ${new URL(request.url).hostname} (datacenter IPs blocked)`);
+      return request;
+    }
+
     // Always try to use a proxy from the central proxy service
     // The service handles bot detection timeouts automatically
     const forceRotation = request.retryCount > 0 || request.metadata.botDetected;