feat(embeddings): best-chunk-per-note on every retrieval surface (#280 step 4)

A note's relevance is now its best chunk's similarity, everywhere:

- semantic_search_notes keeps the indexed raw-distance top-k and over-fetches
  chunk rows (x4, composing with the x3 supersession over-fetch), then
  collapses to first-appearance-per-note — rows arrive distance-ordered, so
  first is best. Every ranked consumer (MCP/REST search, Browse, auto-inject,
  write-path, gate) inherits through the one function.
- list_notes semantic q swaps its join for a correlated MIN-distance
  subquery — the join would have repeated a long note once per matching chunk
  and made total count chunks.
- the duplicate report groups its self-join by note pair on MIN(distance):
  pair similarity = closest chunk pair, and the < join now also drops
  cross-chunk self-pairs that would flag every long note against itself.
- the write gate queries once per chunk of the candidate (capped at 8), so a
  note duplicating an existing record in ONE SECTION is caught — the
  whole-document query diluted exactly the section that mattered.

Integration test now seeds a two-chunk note and pins the collapse against
real pgvector.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UaYUaouG9jjhATyuxCKrQs

src/scribe/services/notes.py

@@ -236,15 +236,25 @@ async def list_notes(
             if query_vec is not None:
                 from scribe.models.embedding import NoteEmbedding
                 from scribe.services.embeddings import INTERACTIVE_SEARCH_THRESHOLD
-                distance = NoteEmbedding.embedding.cosine_distance(query_vec)
-                sem_filter = distance <= (1.0 - INTERACTIVE_SEARCH_THRESHOLD)
-                query = query.join(
-                    NoteEmbedding, NoteEmbedding.note_id == Note.id
-                ).where(sem_filter)
-                count_query = count_query.join(
-                    NoteEmbedding, NoteEmbedding.note_id == Note.id
-                ).where(sem_filter)
-                semantic_order = distance.asc()
+                # Best-chunk-per-note as a correlated MIN, not a join (#280):
+                # a note stores one embedding row PER CHUNK, so the plain join
+                # this used to be would repeat a long note once per matching
+                # chunk — duplicated list rows and a total that counts chunks.
+                # This query is filter-heavy and paginated, never HNSW-bound,
+                # so the scalar subquery costs what the join did.
+                best_distance = (
+                    select(
+                        func.min(
+                            NoteEmbedding.embedding.cosine_distance(query_vec)
+                        )
+                    )
+                    .where(NoteEmbedding.note_id == Note.id)
+                    .scalar_subquery()
+                )
+                sem_filter = best_distance <= (1.0 - INTERACTIVE_SEARCH_THRESHOLD)
+                query = query.where(sem_filter)
+                count_query = count_query.where(sem_filter)
+                semantic_order = best_distance.asc()
             else:
                 terms = _strip_type_nouns(q)
                 for term in terms: