feat(eval): internal-eval stack, judge fixes, and eval web overhaul

- Merge internal-eval into development.sh (single Celery/infra); internal-eval.sh
  wraps with LIFE_ECHO_WITH_INTERNAL_EVAL; EVAL_ATTACH_ONLY for attaching 8001
  when :8000 is already up; document in api/docs/internal-eval.md.
- Evaluation: transcript_for_judge, judge error surfacing, rubric/schema tweaks,
  execution_service and router updates; tests for judge and composite eval.
- Memory: ingest nested transaction for embedding/enrichment rollback safety.
- Conversation WS: logger.exception for pipeline errors (avoid loguru KeyError).
- app-eval-web: Playground saved replays, dialogue turns helper, hash user_id
  for Memoir; Memoir chapter baseline↔DB row compare with title heuristics;
  Stories page (#memoir-stories); Markdown + copy buttons; toolbar/panel UI;
  react-markdown; development proxy and fixture updates.

api/app/features/memory/service.py

@@ -255,35 +255,48 @@ def ingest_transcript_sync(
 
     vectors_written = 0
     embedding_available = False
+    enrichment_ok: bool | None = None
+
     try:
         embedding_provider = get_embedding_provider()
         if embedding_provider is not None:
             embedding_available = embedding_provider.is_available()
-        if chunk_records and embedding_provider is not None:
-            texts = [content for _, content in chunk_records]
-            embeddings = embedding_provider.embed_texts_sync(texts)
-            for (chunk_id, _), emb in zip(chunk_records, embeddings):
-                if emb:
-                    vectors_written += 1
-                    update_chunk_embedding_sync(session, chunk_id, emb)
     except Exception as e:
         logger.warning(
-            "memory embedding 跳过(sync): {} exc_type={}", e, type(e).__name__
+            "memory embedding provider 不可用(sync): {} exc_type={}",
+            e,
+            type(e).__name__,
         )
+        embedding_provider = None
 
-    enrichment_ok: bool | None = None
+    # 向量写入与 enrichment 任一失败时，此前若在**同一事务内**已开始失败的 SQL，
+    # 会导致 session 进入「必须 rollback」状态，进而让后续 commit 抛出
+    # PendingRollbackError，污染 Celery 里共用的 `db`。
+    # 用 SAVEPOINT 包一层：失败仅回滚本段，source/chunks 主体仍可由外层提交。
     try:
-        from app.features.memory.enrichment import enrich_memory_after_ingest_sync
+        with session.begin_nested():
+            if chunk_records and embedding_provider is not None:
+                texts = [content for _, content in chunk_records]
+                embeddings = embedding_provider.embed_texts_sync(texts)
+                for (chunk_id, _), emb in zip(chunk_records, embeddings):
+                    if emb:
+                        vectors_written += 1
+                        update_chunk_embedding_sync(session, chunk_id, emb)
+            if settings.memory_enrichment_enabled:
+                from app.features.memory.enrichment import enrich_memory_after_ingest_sync
 
-        if settings.memory_enrichment_enabled:
-            enrich_memory_after_ingest_sync(session, user_id, source.id, llm=None)
-            enrichment_ok = True
+                enrich_memory_after_ingest_sync(
+                    session, user_id, source.id, llm=None
+                )
+                enrichment_ok = True
     except Exception as e:
+        logger.warning(
+            "memory embedding/enrichment 跳过(sync): {} exc_type={}",
+            e,
+            type(e).__name__,
+        )
         if settings.memory_enrichment_enabled:
             enrichment_ok = False
-        logger.warning(
-            "memory enrichment 跳过(sync): {} exc_type={}", e, type(e).__name__
-        )
 
     session.commit()
     logger.info(