Wip to address https://github.com/crewAIInc/crewAI/issues/1934

src/crewai/agents/crew_agent_executor.py

@@ -3,6 +3,8 @@ import re
 from dataclasses import dataclass
 from typing import Any, Callable, Dict, List, Optional, Union
 
+from litellm.exceptions import AuthenticationError as LiteLLMAuthenticationError
+
 from crewai.agents.agent_builder.base_agent import BaseAgent
 from crewai.agents.agent_builder.base_agent_executor_mixin import CrewAgentExecutorMixin
 from crewai.agents.parser import (
@@ -114,8 +116,10 @@ class CrewAgentExecutor(CrewAgentExecutorMixin):
         """
         formatted_answer = None
         while not isinstance(formatted_answer, AgentFinish):
+            self.iterations += 1
             try:
                 if self._has_reached_max_iterations():
+                    print("Max iterations reached")
                     formatted_answer = self._handle_max_iterations_exceeded(
                         formatted_answer
                     )
@@ -123,8 +127,9 @@ class CrewAgentExecutor(CrewAgentExecutorMixin):
 
                 self._enforce_rpm_limit()
 
+                print("Getting LLM response")
                 answer = self._get_llm_response()
-
+                print(f"LLM response: {answer}")
                 formatted_answer = self._process_llm_response(answer)
 
                 if isinstance(formatted_answer, AgentAction):
@@ -142,15 +147,54 @@ class CrewAgentExecutor(CrewAgentExecutorMixin):
                 formatted_answer = self._handle_output_parser_exception(e)
 
             except Exception as e:
+                print(f"Exception: {e}")
                 if self._is_context_length_exceeded(e):
                     self._handle_context_length()
                     continue
+                elif self._is_litellm_authentication_error(e):
+                    self._handle_litellm_auth_error(e)
+                    break
+                else:
+                    self._printer.print(
+                        content=f"Unhandled exception: {e}",
+                        color="red",
+                    )
 
         self._show_logs(formatted_answer)
         return formatted_answer
 
+    def _is_litellm_authentication_error(self, exception: Exception) -> bool:
+        """Check if the exception is a litellm authentication error."""
+        # Check if the exception is an instance of litellm.AuthenticationError
+        if LiteLLMAuthenticationError and isinstance(
+            exception, LiteLLMAuthenticationError
+        ):
+            return True
+
+        # Alternatively, inspect the exception message
+        error_message = str(exception)
+
+        # Check if 'litellm.AuthenticationError' and 'invalid_api_key' are in the message
+        return "litellm.AuthenticationError" in error_message and (
+            "invalid_api_key" in error_message
+            or "Incorrect API key provided" in error_message
+        )
+
+    def _handle_litellm_auth_error(self, exception: Exception) -> None:
+        """Handle litellm authentication error by informing the user and exiting."""
+        self._printer.print(
+            content="Authentication error with litellm occurred. Please check your API key and configuration.",
+            color="red",
+        )
+        self._printer.print(
+            content=f"Error details: {exception}",
+            color="red",
+        )
+
     def _has_reached_max_iterations(self) -> bool:
         """Check if the maximum number of iterations has been reached."""
+        print(f"Max iterations: {self.max_iter}")
+        print(f"Iterations: {self.iterations}")
         return self.iterations >= self.max_iter
 
     def _enforce_rpm_limit(self) -> None:
@@ -160,10 +204,18 @@ class CrewAgentExecutor(CrewAgentExecutorMixin):
 
     def _get_llm_response(self) -> str:
         """Call the LLM and return the response, handling any invalid responses."""
-        answer = self.llm.call(
+        try:
-            self.messages,
+            answer = self.llm.call(
-            callbacks=self.callbacks,
+                self.messages,
-        )
+                callbacks=self.callbacks,
+            )
+        except Exception as e:
+            self._printer.print(
+                content=f"Error during LLM call: {e}",
+                color="red",
+            )
+            # Re-raise the exception to let _invoke_loop handle it
+            raise e
 
         if not answer:
             self._printer.print(
@@ -184,7 +236,6 @@ class CrewAgentExecutor(CrewAgentExecutorMixin):
                 if FINAL_ANSWER_AND_PARSABLE_ACTION_ERROR_MESSAGE in e.error:
                     answer = answer.split("Observation:")[0].strip()
 
-        self.iterations += 1
         return self._format_answer(answer)
 
     def _handle_agent_action(

src/crewai/utilities/llm_utils.py

@@ -24,12 +24,10 @@ def create_llm(
 
     # 1) If llm_value is already an LLM object, return it directly
     if isinstance(llm_value, LLM):
-        print("LLM value is already an LLM object")
         return llm_value
 
     # 2) If llm_value is a string (model name)
     if isinstance(llm_value, str):
-        print("LLM value is a string")
         try:
             created_llm = LLM(model=llm_value)
             return created_llm
@@ -39,12 +37,10 @@ def create_llm(
 
     # 3) If llm_value is None, parse environment variables or use default
     if llm_value is None:
-        print("LLM value is None")
         return _llm_via_environment_or_fallback()
 
     # 4) Otherwise, attempt to extract relevant attributes from an unknown object
     try:
-        print("LLM value is an unknown object")
         # Extract attributes with explicit types
         model = (
             getattr(llm_value, "model_name", None)