Update app.py

app.py

@@ -91,29 +91,24 @@ def respond(message, history: list[tuple[str, str]], system_message, max_tokens,
 
 
     # Llamar a la API con streaming usando `responses`
-    stream = client.responses.stream(
+    with client.responses.stream(
         model="gpt-5-mini",
         input=messages,
         tools=[{"type": "web_search", "search_context_size": "high"}],
-    )
+    ) as stream:  # ✅ el bloque with maneja correctamente el flujo
+        
+        response = ""
     
-    response = ""
+        for event in stream:  # ✅ ahora sí es iterable dentro del context manager
+            if event.type == "response.output_text.delta":
+                response += event.delta
+                yield response
+            elif event.type == "response.completed":
+                break
     
-    # Recibir los fragmentos del stream
-    for event in stream:
-        if event.type == "response.output_text.delta":
-            response += event.delta
-            yield response
-        elif event.type == "response.completed":
-            break  # 🚨 Importante: terminar cuando el modelo indique fin del stream
-    
-    # Cerrar el stream correctamente
-    stream.close()
-    
-    # Procesar el texto final
-    citations = extract_unique_citations(response)
-    response = response + "\nFuentes: " + citations
-    yield response
+        citations = extract_unique_citations(response)
+        response = response + "\nFuentes: " + citations
+        yield response
 
 # Configuración de la interfaz Gradio
 demo = gr.ChatInterface(