fix(langfuse_otel): clean up review feedback

litellm/integrations/langfuse/langfuse_otel.py

@@ -8,7 +8,6 @@ from litellm._logging import verbose_logger
 from litellm.integrations.arize import _utils
 from litellm.integrations.langfuse.langfuse_otel_attributes import (
     LangfuseLLMObsOTELAttributes,
-    get_langfuse_observation_input_by_type,
 )
 from litellm.integrations.opentelemetry import OpenTelemetry, OpenTelemetryConfig
 from litellm.types.integrations.langfuse_otel import (
@@ -244,14 +243,6 @@ class LangfuseOtelLogger(OpenTelemetry):
         metadata = LangfuseOtelLogger._extract_langfuse_metadata(kwargs)
         LangfuseOtelLogger._set_metadata_attributes(span=span, metadata=metadata)
 
-        input_payload = get_langfuse_observation_input_by_type(kwargs)
-        if input_payload is not None:
-            safe_set_attribute(
-                span,
-                LangfuseSpanAttributes.OBSERVATION_INPUT.value,
-                safe_dumps(input_payload),
-            )
-
         LangfuseOtelLogger._set_observation_output(span=span, response_obj=response_obj)
 
     @staticmethod

litellm/integrations/langfuse/langfuse_otel_attributes.py

@@ -128,7 +128,9 @@ def get_langfuse_observation_input_by_type(
         return prompt
 
     prompt: dict[str, Any] = {}
-    if "messages" in kwargs:
+    # Preserve explicit empty chat message lists without serializing `null`
+    # when callers pass `messages=None`.
+    if messages is not None:
         prompt["messages"] = messages
     for key in ("functions", "tools"):
         value = optional_params.get(key)

tests/test_litellm/integrations/test_langfuse_otel.py

@@ -257,21 +257,14 @@ class TestLangfuseOtelIntegration:
             ],
         )
 
-        kwargs = {
-            "messages": [{"role": "user", "content": "What's the weather in Tokyo?"}],
-        }
-
         with patch(
             "litellm.integrations.arize._utils.safe_set_attribute"
         ) as mock_safe_set_attribute:
             LangfuseOtelLogger._set_langfuse_specific_attributes(
-                MagicMock(), kwargs, response_obj
+                MagicMock(), {}, response_obj
             )
 
             expect_output = {
-                LangfuseSpanAttributes.OBSERVATION_INPUT.value: [
-                    {"role": "user", "content": "What's the weather in Tokyo?"}
-                ],
                 LangfuseSpanAttributes.OBSERVATION_OUTPUT.value: {
                     "role": "assistant",
                     "content": "The weather in Tokyo is sunny.",
@@ -572,6 +565,46 @@ class TestLangfuseOtelResponsesAPI:
             in actual_attributes["langfuse.observation.input"]
         )
 
+    def test_langfuse_observation_input_skips_null_messages(self):
+        """Explicit null messages should not serialize as `{\"messages\": null}`."""
+        kwargs = {
+            "messages": None,
+            "optional_params": {
+                "tools": [{"type": "function", "name": "list_applications"}]
+            },
+        }
+
+        from litellm.integrations.langfuse.langfuse_otel_attributes import (
+            get_langfuse_observation_input_by_type,
+        )
+
+        assert get_langfuse_observation_input_by_type(kwargs) == {
+            "tools": [{"type": "function", "name": "list_applications"}]
+        }
+
+    def test_langfuse_specific_attributes_do_not_rewrite_observation_input(self):
+        """Observation input should be owned by set_messages, not rewritten later."""
+        from litellm.types.integrations.langfuse_otel import LangfuseSpanAttributes
+
+        kwargs = {
+            "input": [{"role": "user", "content": "hello"}],
+            "optional_params": {"instructions": "Reply briefly."},
+            "litellm_params": {"metadata": {"generation_name": "gen-name"}},
+        }
+
+        with patch(
+            "litellm.integrations.arize._utils.safe_set_attribute"
+        ) as mock_safe_set_attribute:
+            LangfuseOtelLogger._set_langfuse_specific_attributes(
+                MagicMock(), kwargs, None
+            )
+
+            actual_keys = [
+                call.args[1] for call in mock_safe_set_attribute.call_args_list
+            ]
+            assert LangfuseSpanAttributes.GENERATION_NAME.value in actual_keys
+            assert LangfuseSpanAttributes.OBSERVATION_INPUT.value not in actual_keys
+
     def test_responses_api_metadata_extraction(self):
         """Test that metadata is correctly extracted from ResponsesAPI kwargs."""
         # Clean up any existing module mocks