From 36e064b240bca43981df3d5ee17be59c177eb861 Mon Sep 17 00:00:00 2001
From: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
Date: Tue, 11 Mar 2025 07:33:05 +0000
Subject: [PATCH] Address PR feedback: Enhance type safety and add tests for
 multimodal validation

Co-Authored-By: Joe Moura <joao@crewai.com>
---
 src/crewai/utilities/events/llm_events.py | 40 +++++++++++++++++++++--
 tests/test_multimodal_validation.py       | 24 ++++++++++++++
 2 files changed, 62 insertions(+), 2 deletions(-)

diff --git a/src/crewai/utilities/events/llm_events.py b/src/crewai/utilities/events/llm_events.py
index 10a648e86..2f2881e3e 100644
--- a/src/crewai/utilities/events/llm_events.py
+++ b/src/crewai/utilities/events/llm_events.py
@@ -1,5 +1,5 @@
 from enum import Enum
-from typing import Any, Dict, List, Optional, Union
+from typing import Any, Dict, List, Literal, Optional, Union
 
 from crewai.utilities.events.base_events import CrewEvent
 
@@ -11,11 +11,47 @@ class LLMCallType(Enum):
     LLM_CALL = "llm_call"
 
 
+class ContentType(str, Enum):
+    """Types of content in multimodal messages"""
+    
+    TEXT = "text"
+    IMAGE_URL = "image_url"
+
+
 class LLMCallStartedEvent(CrewEvent):
     """Event emitted when a LLM call starts"""
 
     type: str = "llm_call_started"
-    messages: Union[str, List[Dict[str, Any]]]
+    messages: Union[
+        str,
+        List[Union[
+            str,
+            Dict[str, Union[
+                str,
+                List[Dict[str, Union[
+                    str,
+                    Dict[Literal["url"], str]
+                ]]]
+            ]]
+        ]]
+    ]
+    """
+    Supports both string messages and structured messages including multimodal content.
+    Formats supported:
+    1. Simple string: "This is a message"
+    2. List of message objects: [{"role": "user", "content": "Hello"}]
+    3. Mixed list with strings and objects: ["Simple message", {"role": "user", "content": "Hello"}]
+    4. Multimodal format:
+    {
+        'role': str,
+        'content': List[
+            Union[
+                Dict[Literal["type", "text"], str],
+                Dict[Literal["type", "image_url"], Dict[str, str]]
+            ]
+        ]
+    }
+    """
     tools: Optional[List[dict]] = None
     callbacks: Optional[List[Any]] = None
     available_functions: Optional[Dict[str, Any]] = None
diff --git a/tests/test_multimodal_validation.py b/tests/test_multimodal_validation.py
index 9057e4456..536d0f5d5 100644
--- a/tests/test_multimodal_validation.py
+++ b/tests/test_multimodal_validation.py
@@ -62,3 +62,27 @@ def test_llm_call_started_event_with_standard_messages():
     assert event.messages[0]['content'] == 'You are a helpful assistant'
     assert event.messages[1]['role'] == 'user'
     assert event.messages[1]['content'] == 'Hello, how are you?'
+
+
+def test_llm_call_started_event_with_mixed_content():
+    """Test that LLMCallStartedEvent handles mixed content types."""
+    mixed_messages = [
+        "Simple string message",
+        {
+            'role': 'user',
+            'content': [
+                {'type': 'text', 'text': 'With image'},
+                {'type': 'image_url', 'image_url': {'url': 'https://example.com/image.jpg'}},
+            ],
+        }
+    ]
+    
+    # This should not raise a ValidationError
+    event = LLMCallStartedEvent(messages=mixed_messages)
+    
+    # Verify the event was created correctly
+    assert isinstance(event.messages, list)
+    assert isinstance(event.messages[0], str)
+    assert isinstance(event.messages[1], dict)
+    assert event.messages[1]['role'] == 'user'
+    assert isinstance(event.messages[1]['content'], list)