[OPIK-1779] enable python metrics for threads (#2741)

* [OPIK-1779] enable python metrics for threads

* [OPIK-1779] update python thread prompt

* [OPIK-1779] add comment to python thread metric

---------

Co-authored-by: Yaroslav Boiko <yaroslavboiko@ip-192-168-1-15.ec2.internal>

Author: awkoy

apps/opik-frontend/src/components/pages-shared/automations/AddEditRuleDialog/AddEditRuleDialog.tsx

@@ -183,24 +183,20 @@ const AddEditRuleDialog: React.FC<AddEditRuleDialogProps> = ({
   const handleScopeChange = useCallback(
     (value: EVALUATORS_RULE_SCOPE) => {
       const applyChange = () => {
-        if (value === EVALUATORS_RULE_SCOPE.thread) {
-          form.setValue("uiType", UI_EVALUATORS_RULE_TYPE.llm_judge);
-        }
-
         const { uiType } = form.getValues();
         const type = getBackendRuleType(value, uiType);
 
         form.setValue("scope", value);
         form.setValue("type", type);
 
-        form.resetField("llmJudgeDetails", {
-          defaultValue: cloneDeep(DEFAULT_LLM_AS_JUDGE_DATA[value]),
-          keepDirty: false,
-        });
-        form.resetField("pythonCodeDetails", {
-          defaultValue: cloneDeep(DEFAULT_PYTHON_CODE_DATA[value]),
-          keepDirty: false,
-        });
+        form.setValue(
+          "llmJudgeDetails",
+          cloneDeep(DEFAULT_LLM_AS_JUDGE_DATA[value]),
+        );
+        form.setValue(
+          "pythonCodeDetails",
+          cloneDeep(DEFAULT_PYTHON_CODE_DATA[value]),
+        );
       };
 
       if (
@@ -464,7 +460,7 @@ const AddEditRuleDialog: React.FC<AddEditRuleDialogProps> = ({
                   )}
                 />
 
-                {!isEdit && !isThreadScope && (
+                {!isEdit && (
                   <FormField
                     control={form.control}
                     name="uiType"

apps/opik-frontend/src/constants/llm.ts

@@ -585,14 +585,23 @@ export const DEFAULT_PYTHON_CODE_THREAD_DATA: PythonCodeDetailsThreadForm = {
   metric:
     "from typing import Union, List, Any\n" +
     "from opik.evaluation.metrics import base_metric, score_result\n" +
-    "from opik.evaluation.types import Conversation\n" +
+    "from opik.evaluation.metrics.conversation import conversation_thread_metric, types\n" +
     "\n" +
-    "class MyCustomMetric(base_metric.BaseMetric):\n" +
+    "class MyCustomMetric(conversation_thread_metric.ConversationThreadMetric):\n" +
     '    """A custom metric for evaluating conversation threads."""\n' +
+    "    def __init__(\n" +
+    "        self,\n" +
+    '        name: str = "my_custom_thread_metric",\n' +
+    "    ):\n" +
+    "        super().__init__(\n" +
+    "            name=name,\n" +
+    "        )\n" +
     "\n" +
     "    def score(\n" +
-    "        self, conversation: Conversation, **kwargs: Any\n" +
+    "        self, conversation: types.Conversation, **kwargs: Any\n" +
     "    ) -> Union[score_result.ScoreResult, List[score_result.ScoreResult]]:\n" +
+    "        # conversation is a List[Dict] where each dict has:\n" +
+    '        # {"role": "user" | "assistant", "content": "message text"}\n' +
     "        # Add you logic here\n" +
     "\n" +
     "        return score_result.ScoreResult(\n" +