diff --git a/llama_stack/providers/utils/telemetry/dataset_mixin.py b/llama_stack/providers/utils/telemetry/dataset_mixin.py
index c52ec7d89..d67462b05 100644
--- a/llama_stack/providers/utils/telemetry/dataset_mixin.py
+++ b/llama_stack/providers/utils/telemetry/dataset_mixin.py
@@ -49,7 +49,7 @@ class TelemetryDatasetMixin:
         attributes_to_return: List[str],
         max_depth: Optional[int] = None,
     ) -> QuerySpansResponse:
-        traces = await self.query_traces(attribute_filters=attribute_filters)
+        traces = await self.query_traces(attribute_filters=attribute_filters).data
         spans = []
 
         for trace in traces:
diff --git a/tests/client-sdk/agents/test_agents.py b/tests/client-sdk/agents/test_agents.py
index 747b64dd1..dd6a4189e 100644
--- a/tests/client-sdk/agents/test_agents.py
+++ b/tests/client-sdk/agents/test_agents.py
@@ -178,6 +178,14 @@ def test_builtin_tool_web_search(llama_stack_client, agent_config):
     assert "Tool:brave_search Response:" in logs_str
     assert "mark zuckerberg" in logs_str.lower()
     assert "No Violation" in logs_str
+    spans = llama_stack_client.telemetry.query_spans(
+        attribute_filters=[
+            {"key": "session_id", "op": "eq", "value": session_id},
+        ],
+        attributes_to_return=["input", "output"],
+    )
+    breakpoint()
+    assert len(spans) > 1
 
 
 def test_builtin_tool_code_execution(llama_stack_client, agent_config):