Fix cluster client communication architecture

- Modified cluster client to connect to backend's TCP web port instead of worker Unix socket
- Backend acts as proper bridge: web interface (TCP) ↔ workers (Unix socket)
- Cluster client now communicates with backend the same way as web interface
- This fixes the timeout issue and ensures proper job flow through the backend

vidai/cluster_client.py

@@ -299,13 +299,13 @@ class ClusterClient:
 
         print(f"Received job assignment: {job_id} for process type: {process_type}")
 
-        # Forward job to local backend, which will route it to the appropriate worker
+        # Forward job to local backend via TCP (web interface), which will route it to the appropriate worker
         try:
             from .comm import SocketCommunicator
-            from .compat import get_socket_path
+            from .config import get_backend_web_port
 
-            # Connect to local backend
-            backend_comm = SocketCommunicator(socket_path=get_socket_path('worker'), comm_type='unix')
+            # Connect to local backend web server
+            backend_comm = SocketCommunicator(host='localhost', port=get_backend_web_port(), comm_type='tcp')
             backend_comm.connect()
 
             # Send job to backend
@@ -333,12 +333,12 @@ class ClusterClient:
         """Monitor for job result from local backend."""
         try:
             from .comm import SocketCommunicator
-            from .compat import get_socket_path
+            from .config import get_backend_web_port
 
             # Poll for result
             for _ in range(300):  # Poll for up to 5 minutes (300 * 1s)
                 try:
-                    backend_comm = SocketCommunicator(socket_path=get_socket_path('worker'), comm_type='unix')
+                    backend_comm = SocketCommunicator(host='localhost', port=get_backend_web_port(), comm_type='tcp')
                     backend_comm.connect()
 
                     # Send get_result request