Python: Fix IndexError when reasoning models produce reasoning-only messages in Magentic-One workflow

python/packages/core/agent_framework/openai/_chat_client.py

@@ -548,6 +548,7 @@ def _prepare_message_for_openai(self, message: Message) -> list[dict[str, Any]]:
             return []
 
         all_messages: list[dict[str, Any]] = []
+        pending_reasoning: Any = None
         for content in message.contents:
             # Skip approval content - it's internal framework state, not for the LLM
             if content.type in ("function_approval_request", "function_approval_response"):
@@ -575,15 +576,33 @@ def _prepare_message_for_openai(self, message: Message) -> list[dict[str, Any]]:
                     # Functions returning None should still have a tool result message
                     args["content"] = content.result if content.result is not None else ""
                 case "text_reasoning" if (protected_data := content.protected_data) is not None:
-                    all_messages[-1]["reasoning_details"] = json.loads(protected_data)
+                    # Buffer reasoning to attach to the next message with content/tool_calls
+                    pending_reasoning = json.loads(protected_data)
                 case _:
                     if "content" not in args:
                         args["content"] = []
                     # this is a list to allow multi-modal content
                     args["content"].append(self._prepare_content_for_openai(content))  # type: ignore
             if "content" in args or "tool_calls" in args:
+                if pending_reasoning is not None:
+                    args["reasoning_details"] = pending_reasoning
+                    pending_reasoning = None
                 all_messages.append(args)
 
+        # If reasoning was the only content, emit a valid message with empty content
+        if pending_reasoning is not None:
+            if all_messages:
+                all_messages[-1]["reasoning_details"] = pending_reasoning
+            else:
+                pending_args: dict[str, Any] = {
+                    "role": message.role,
+                    "content": "",
+                    "reasoning_details": pending_reasoning,
+                }
+                if message.author_name and message.role != "tool":
+                    pending_args["name"] = message.author_name
+                all_messages.append(pending_args)
+
         # Flatten text-only content lists to plain strings for broader
         # compatibility with OpenAI-like endpoints (e.g. Foundry Local).
         # See https://github.com/microsoft/agent-framework/issues/4084

python/packages/core/tests/openai/test_openai_chat_client.py

@@ -643,6 +643,110 @@ def test_prepare_message_with_text_reasoning_content(openai_unit_test_env: dict[
     assert prepared[0]["content"] == "The answer is 42."
 
 
+def test_prepare_message_with_only_text_reasoning_content(openai_unit_test_env: dict[str, str]) -> None:
+    """Test that a message with only text_reasoning content does not raise IndexError.
+
+    Regression test for https://github.com/microsoft/agent-framework/issues/4384
+    Reasoning models (e.g. gpt-5-mini) may produce reasoning_details without text content,
+    which previously caused an IndexError when preparing messages.
+    """
+    client = OpenAIChatClient()
+
+    mock_reasoning_data = {
+        "effort": "high",
+        "summary": "Deep analysis of the problem",
+    }
+
+    reasoning_content = Content.from_text_reasoning(text=None, protected_data=json.dumps(mock_reasoning_data))
+
+    # Message with only reasoning content and no text
+    message = Message(
+        role="assistant",
+        contents=[reasoning_content],
+    )
+
+    prepared = client._prepare_message_for_openai(message)
+
+    # Should have one message with reasoning_details
+    assert len(prepared) == 1
+    assert prepared[0]["role"] == "assistant"
+    assert "reasoning_details" in prepared[0]
+    assert prepared[0]["reasoning_details"] == mock_reasoning_data
+    # Message should also include a content field to be a valid Chat Completions payload
+    assert "content" in prepared[0]
+    assert prepared[0]["content"] == ""
+
+
+def test_prepare_message_with_text_reasoning_before_text(openai_unit_test_env: dict[str, str]) -> None:
+    """Test that text_reasoning content appearing before text content is handled correctly.
+
+    Regression test for https://github.com/microsoft/agent-framework/issues/4384
+    """
+    client = OpenAIChatClient()
+
+    mock_reasoning_data = {
+        "effort": "medium",
+        "summary": "Quick analysis",
+    }
+
+    reasoning_content = Content.from_text_reasoning(text=None, protected_data=json.dumps(mock_reasoning_data))
+
+    # Reasoning appears before text content
+    message = Message(
+        role="assistant",
+        contents=[
+            reasoning_content,
+            Content.from_text(text="The answer is 42."),
+        ],
+    )
+
+    prepared = client._prepare_message_for_openai(message)
+
+    # Should produce exactly one message without raising IndexError
+    assert len(prepared) == 1
+
+    # Reasoning details should be present on the message
+    assert "reasoning_details" in prepared[0]
+    assert prepared[0]["reasoning_details"] == mock_reasoning_data
+    assert prepared[0]["content"] == "The answer is 42."
+
+
+def test_prepare_message_with_text_reasoning_before_function_call(openai_unit_test_env: dict[str, str]) -> None:
+    """Test that text_reasoning content appearing before a function call is handled correctly.
+
+    Regression test for https://github.com/microsoft/agent-framework/issues/4384
+    """
+    client = OpenAIChatClient()
+
+    mock_reasoning_data = {
+        "effort": "medium",
+        "summary": "Deciding to call a function",
+    }
+
+    reasoning_content = Content.from_text_reasoning(text=None, protected_data=json.dumps(mock_reasoning_data))
+
+    # Reasoning appears before function call content
+    message = Message(
+        role="assistant",
+        contents=[
+            reasoning_content,
+            Content.from_function_call(call_id="call_abc", name="get_weather", arguments='{"city": "Seattle"}'),
+        ],
+    )
+
+    prepared = client._prepare_message_for_openai(message)
+
+    # Should produce exactly one message
+    assert len(prepared) == 1
+
+    # The message should carry the reasoning details and tool_calls
+    assert "reasoning_details" in prepared[0]
+    assert prepared[0]["reasoning_details"] == mock_reasoning_data
+    assert "tool_calls" in prepared[0]
+    assert prepared[0]["tool_calls"][0]["function"]["name"] == "get_weather"
+    assert prepared[0]["role"] == "assistant"
+
+
 def test_function_approval_content_is_skipped_in_preparation(openai_unit_test_env: dict[str, str]) -> None:
     """Test that function approval request and response content are skipped."""
     client = OpenAIChatClient()