try fix llm with retry

Author: neagualexa

src/agents/llm_factory.py

@@ -79,8 +79,13 @@ def __init__(self, temperature: int = 0):
             model=os.environ['GOOGLE_AI_MODEL'],
             temperature=temperature,
             google_api_key=os.environ['GOOGLE_AI_API_KEY'],
-            max_retries=0, 
+        )
+
+        #    This wrapper will catch errors and retry the call itself.
+        #    It does NOT pass 'max_retries' to the underlying model.
+        self._google_llm_with_retries = self._google_llm.with_retry(
+            stop_after_attempt=1,  # Corresponds to max_retries
         )
 
     def get_llm(self):
-        return self._google_llm
+        return self._google_llm_with_retries