feat(api)!: memory single chain — async MemoryService, strict eval closure

Route all memory ingest/retrieve/enrichment/compaction through async MemoryService.
Remove legacy sync memory implementations (ingest/retrieve/compaction); Celery and
memoir Phase2 call asyncio.run into MemoryService-backed helpers.

Memoir Phase1 batch ingest uses MemoryService.ingest_transcripts_batch; drop chapters.
evidence_bundle_json mirror (Alembic 0015). Evaluation uses snapshot/link-only bundles;
raise EvidenceClosureMissing instead of partial/fallback lineage tiers.

Split memoir state into NarrativeCoverageState and InterviewControlState; delete the
_interview_meta_store adapter layer. Remove rolling-query and recent-fact fallback
settings from config and evidence assembly.

Update judges, docs, tests, and PlaygroundPage alignment.

Made-with: Cursor

api/tests/test_memory_evidence.py

@@ -2,73 +2,18 @@
 
 import pytest
 
-from app.core.config import settings
 from app.features.memory import evidence as evidence_mod
-from app.features.memory.evidence_format import format_evidence_chunks_for_chat_prompt
 from app.features.memory.evidence import (
     EMPTY_EVIDENCE_BUNDLE,
     _facts_to_dicts,
     _stories_to_dicts,
     _timeline_to_dicts,
     retrieve_evidence_bundle_async,
-    retrieve_evidence_bundle_sync,
 )
+from app.features.memory.evidence_format import format_evidence_chunks_for_chat_prompt
 from app.features.memory.schemas import EvidenceBundle
 
 
-class _FakeEmbedding:
-    def is_available(self) -> bool:
-        return True
-
-    def embed_text_sync(self, text: str) -> list[float]:
-        return [0.25, 0.5, 0.75]
-
-
-def test_retrieve_evidence_bundle_sync_uses_vector_search(
-    monkeypatch: pytest.MonkeyPatch,
-) -> None:
-    searched: list[tuple] = []
-
-    def fake_search(session, user_id, emb, top_k):
-        searched.append((user_id, emb, top_k))
-        return [
-            {
-                "id": "c1",
-                "content": "chunk body",
-                "chunk_index": 0,
-                "distance": 0.1,
-            }
-        ]
-
-    def fake_meta(user_id, q, top_k):
-        return {
-            "relevant_facts": [],
-            "timeline_hints": [],
-            "relevant_summaries": [],
-            "relevant_stories": [],
-        }
-
-    monkeypatch.setattr(evidence_mod, "search_chunks_vector_sync", fake_search)
-    monkeypatch.setattr(
-        evidence_mod, "fetch_evidence_metadata_parallel_sync", fake_meta
-    )
-
-    out = retrieve_evidence_bundle_sync(
-        session=object(),
-        user_id="u1",
-        query="  hello  ",
-        top_k=7,
-        embedding_provider=_FakeEmbedding(),
-    )
-    assert len(searched) == 1
-    assert searched[0][0] == "u1"
-    assert searched[0][1] == [0.25, 0.5, 0.75]
-    assert searched[0][2] == 7
-    assert out["relevant_chunks"] == [
-        {"id": "c1", "content": "chunk body", "chunk_index": 0},
-    ]
-
-
 def test_empty_evidence_bundle_keys() -> None:
     assert set(EMPTY_EVIDENCE_BUNDLE.keys()) == {
         "relevant_chunks",
@@ -238,10 +183,7 @@ async def test_retrieve_evidence_bundle_async_non_empty_merges_precomputed_chunk
     assert out == {"relevant_chunks": merged, **meta}
 
 
-async def test_empty_query_evidence_bundle_async_and_sync_aligned_when_rolling_off(
-    monkeypatch: pytest.MonkeyPatch,
-) -> None:
-    monkeypatch.setattr(settings, "memory_evidence_empty_query_include_rolling", False)
+async def test_empty_query_evidence_bundle_async_returns_empty() -> None:
     out_a = await retrieve_evidence_bundle_async(
         object(),
         "u1",
@@ -250,11 +192,3 @@ async def test_empty_query_evidence_bundle_async_and_sync_aligned_when_rolling_o
         merged_chunk_dicts=[],
     )
     assert out_a == dict(EMPTY_EVIDENCE_BUNDLE)
-    out_s = retrieve_evidence_bundle_sync(
-        session=object(),
-        user_id="u1",
-        query="",
-        top_k=10,
-        embedding_provider=None,
-    )
-    assert out_s == dict(EMPTY_EVIDENCE_BUNDLE)