Apply suggestions from code review

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

python/packages/core/agent_framework/_evaluation.py

@@ -1732,6 +1732,8 @@ async def evaluate_workflow(
     if workflow_result is None and queries is None:
         raise ValueError("Provide either 'workflow_result' or 'queries'.")
 
+    if expected_output is not None and queries is None:
+        raise ValueError("Provide 'queries' when using 'expected_output'; 'expected_output' is not supported with 'workflow_result' only.")
     if expected_output is not None and queries is not None and len(expected_output) != len(queries):
         raise ValueError(f"Got {len(queries)} queries but {len(expected_output)} expected_output values.")
 

python/packages/foundry/agent_framework_foundry/_foundry_evals.py

@@ -692,7 +692,7 @@ class FoundryEvals:
                 ]
             if item.context:
                 d["context"] = item.context
-            if item.expected_output:
+            if item.expected_output is not None:
                 d["ground_truth"] = item.expected_output
             dicts.append(d)
 

python/samples/05-end-to-end/evaluation/foundry_evals/evaluate_workflow_sample.py

@@ -5,7 +5,7 @@
 This sample demonstrates three patterns:
 1. Post-hoc: Run the workflow, then evaluate the result you already have.
 2. Run + evaluate: Pass queries and let evaluate_workflow() run the workflow for you.
-3. Similarity: Evaluate an agent's output against ground-truth reference answers.
+3. Similarity: Evaluate the workflow's final output against ground-truth reference answers.
 
 Patterns 1 & 2 return a list of results (one per provider), each with a per-agent
 breakdown in sub_results so you can identify which agent is underperforming.