feat(evaluation): memoir readiness, judge/replay updates, eval web playground

Add memoir_readiness_service and router tests; extend judge schemas/services, replay_service, and conversation rubric; align story route agent, payload, prompts, and story_pipeline_sync; update agent logging, config, and DI. Document internal-eval; add replayDraft util and PlaygroundPage changes in app-eval-web.

api/app/features/evaluation/replay_service.py

@@ -98,7 +98,7 @@ class ReplayConversationService:
         fixture_filename: str,
         flush_memoir_after: bool,
         skip_tts: bool,
-    ) -> tuple[int, list[str]]:
+    ) -> tuple[int, list[str], list[str]]:
         try:
             turns, _ = read_user_export_fixture(fixture_filename)
         except ValueError as e:
@@ -108,13 +108,13 @@ class ReplayConversationService:
         utterances = [u.strip() for u, _ in turns if (u or "").strip()]
         if not utterances:
             raise EvaluationBadRequestError("fixture produced no user utterances")
-        n = await self.replay_utterances(
+        n, segment_ids = await self.replay_utterances(
             conversation_id=conversation_id,
             utterances=utterances,
             flush_memoir_after=flush_memoir_after,
             skip_tts=skip_tts,
         )
-        return n, utterances
+        return n, utterances, segment_ids
 
     async def replay_utterances(
         self,
@@ -123,7 +123,7 @@ class ReplayConversationService:
         utterances: list[str],
         flush_memoir_after: bool,
         skip_tts: bool,
-    ) -> int:
+    ) -> tuple[int, list[str]]:
         cid = (conversation_id or "").strip()
         if not cid:
             raise EvaluationBadRequestError("conversation_id is required")
@@ -136,11 +136,13 @@ class ReplayConversationService:
 
         conv_service = ConversationService(self._db, self._quota)
         count = 0
+        segment_ids: list[str] = []
         for raw in utterances:
             text = (raw or "").strip()
             if not text:
                 continue
             segment = await conv_service.create_user_segment(conv, conv.user_id, text)
+            segment_ids.append(segment.id)
             ts = segment.created_at or conv.last_message_at
             await background_runner.queue_message(
                 conv.user_id,
@@ -169,4 +171,4 @@ class ReplayConversationService:
             flush_memoir_after,
             skip_tts,
         )
-        return count
+        return count, segment_ids