Strip nulls from remote eval prompt parameter defaults

 ## Summary

  Fix Python remote eval prompt parameter defaults that include explicit `None` / JSON `null` values from `PromptData.as_dict()`.

  `PromptData.as_dict()` currently preserves optional prompt fields like `tools`, `name`, `function_call`, and `tool_calls` as `None`. When these values are used as a `type:
  "prompt"` parameter default, the remote eval manifest contains explicit `null`s. The playground schema expects these fields to be omitted when unset, so the remote eval can fail
  to appear or fail to parse.

  This strips `None` values recursively only when serializing prompt parameter defaults, leaving the generic `PromptData.as_dict()` behavior unchanged.

  ## Test Plan

  - Add/updated Python parameter serialization test covering a `PromptData` prompt default with unset optional fields.
  - Verify serialized remote eval parameter container omits `tools`, `name`, `function_call`, and `tool_calls` when unset.
  - Manually verify `bt eval <repro>.py --dev` lists the remote eval in the playground.

Author: evanmkeith

py/src/braintrust/parameters.py

@@ -215,14 +215,32 @@ def is_eval_parameter_schema(schema: Any) -> bool:
     return True
 
 
+def _strip_none_values(value: Any) -> Any:
+    """Recursively drop dict entries whose value is ``None``.
+
+    Prompt parameter defaults are serialized from Python prompt dataclasses,
+    which preserve explicit ``None`` values for optional fields (e.g. a message's
+    ``name``/``function_call``/``tool_calls`` or a chat block's ``tools``). The
+    JS/UI evaluator manifest schema treats those fields as optional-when-absent
+    rather than nullable, so an explicit ``null`` fails validation and the remote
+    eval never appears in the playground. Stripping ``None`` keeps the emitted
+    default aligned with what the UI schema expects.
+    """
+    if isinstance(value, dict):
+        return {key: _strip_none_values(item) for key, item in value.items() if item is not None}
+    if isinstance(value, list):
+        return [_strip_none_values(item) for item in value]
+    return value
+
+
 def _prompt_data_to_dict(
     prompt_data: PromptDataDict | PromptData | None,
 ) -> dict[str, Any] | None:
     if prompt_data is None:
         return None
     if isinstance(prompt_data, PromptData):
-        return prompt_data.as_dict()
-    return dict(prompt_data)
+        return _strip_none_values(prompt_data.as_dict())
+    return _strip_none_values(dict(prompt_data))
 
 
 def _create_prompt(name: str, prompt_data: dict[str, Any]) -> "Prompt":

py/src/braintrust/test_parameters.py

@@ -5,8 +5,10 @@
     RemoteEvalParameters,
     parameters_to_json_schema,
     serialize_eval_parameters,
+    serialize_remote_eval_parameters_container,
     validate_parameters,
 )
+from braintrust.prompt import PromptChatBlock, PromptData, PromptMessage
 
 
 HAS_PYDANTIC = importlib.util.find_spec("pydantic") is not None
@@ -563,3 +565,73 @@ def test_parameters_to_json_schema_does_not_mark_passthrough_values_required():
     )
 
     assert "required" not in schema
+
+
+def _assert_no_none_values(node):
+    if isinstance(node, dict):
+        for key, value in node.items():
+            assert value is not None, f"unexpected None at key {key!r}"
+            _assert_no_none_values(value)
+    elif isinstance(node, list):
+        for item in node:
+            _assert_no_none_values(item)
+
+
+def test_prompt_parameter_defaults_omit_none_values():
+    prompt_data = PromptData(
+        prompt=PromptChatBlock(messages=[PromptMessage(role="user", content="{{input}}")]),
+        options={"model": "gpt-5-mini"},
+    )
+
+    serialized = serialize_remote_eval_parameters_container(
+        {
+            "grouping_prompt": {
+                "type": "prompt",
+                "description": "Grouping prompt",
+                "default": prompt_data,
+            }
+        }
+    )
+
+    default = serialized["schema"]["grouping_prompt"]["default"]
+    assert "tools" not in default["prompt"]
+    assert "name" not in default["prompt"]["messages"][0]
+    assert "function_call" not in default["prompt"]["messages"][0]
+    assert "tool_calls" not in default["prompt"]["messages"][0]
+    _assert_no_none_values(default)
+
+
+def test_prompt_parameter_defaults_omit_none_values_from_dict():
+    prompt_default = {
+        "prompt": {
+            "type": "chat",
+            "messages": [
+                {
+                    "role": "user",
+                    "content": "{{input}}",
+                    "name": None,
+                    "function_call": None,
+                    "tool_calls": None,
+                }
+            ],
+            "tools": None,
+        },
+        "options": {"model": "gpt-5-mini"},
+    }
+
+    serialized = serialize_eval_parameters(
+        {
+            "grouping_prompt": {
+                "type": "prompt",
+                "description": "Grouping prompt",
+                "default": prompt_default,
+            }
+        }
+    )
+
+    default = serialized["grouping_prompt"]["default"]
+    assert "tools" not in default["prompt"]
+    assert "name" not in default["prompt"]["messages"][0]
+    assert "function_call" not in default["prompt"]["messages"][0]
+    assert "tool_calls" not in default["prompt"]["messages"][0]
+    _assert_no_none_values(default)