fix: (search) oversample to reduce duplicate errors

src/crud/message.py

@@ -1,3 +1,4 @@
+from collections.abc import Sequence
 from datetime import datetime
 from logging import getLogger
 from typing import Any
@@ -18,6 +19,21 @@ from .session import get_or_create_session
 logger = getLogger(__name__)
 
 
+def _deduplicate_messages(
+    messages: Sequence[models.Message], limit: int
+) -> list[models.Message]:
+    """Deduplicate messages by public_id, preserving input order."""
+    seen: set[str] = set()
+    result: list[models.Message] = []
+    for msg in messages:
+        if msg.public_id not in seen:
+            seen.add(msg.public_id)
+            result.append(msg)
+            if len(result) >= limit:
+                break
+    return result
+
+
 def _apply_token_limit(
     base_conditions: list[ColumnElement[Any]], token_limit: int
 ) -> Select[tuple[models.Message]]:
@@ -606,11 +622,15 @@ async def _search_messages_external(
     if session_name:
         vector_filters["session_name"] = session_name
 
-    # Oversample: chunks can map to the same message, so fetch extra
+    # Oversample: chunks can map to the same message, and date filters are
+    # applied post-fetch (vector stores don't support temporal filtering),
+    # so fetch extra to compensate for both deduplication and filtering.
+    has_date_filters = after_date is not None or before_date is not None
+    oversample = 6 if has_date_filters else 3
     vector_results = await external_vector_store.query(
         namespace,
         query_embedding,
-        top_k=limit * 3,
+        top_k=limit * oversample,
         filters=vector_filters if vector_filters else None,
     )
 
@@ -682,6 +702,9 @@ async def search_messages(
 
     if settings.VECTOR_STORE.TYPE == "pgvector" or not settings.VECTOR_STORE.MIGRATED:
         # pgvector path: cosine distance in SQL
+        # Oversample because a message with multiple embedding chunks can
+        # produce duplicate rows; we deduplicate in Python to preserve HNSW
+        # index usage (a DISTINCT ON subquery would prevent the index scan).
         match_stmt = (
             select(models.Message)
             .join(
@@ -692,7 +715,7 @@ async def search_messages(
             .order_by(
                 models.MessageEmbedding.embedding.cosine_distance(query_embedding)
             )
-            .limit(limit)
+            .limit(limit * 2)
         )
 
         if session_name:
@@ -701,7 +724,7 @@ async def search_messages(
             )
 
         result = await db.execute(match_stmt)
-        matched_messages = list(result.scalars().all())
+        matched_messages = _deduplicate_messages(result.scalars().all(), limit)
     else:
         # External vector store path
         matched_messages = await _search_messages_external(
@@ -845,6 +868,7 @@ async def search_messages_temporal(
 
     if settings.VECTOR_STORE.TYPE == "pgvector" or not settings.VECTOR_STORE.MIGRATED:
         # pgvector path: cosine distance in SQL with date filters
+        # Oversample to handle chunk duplicates (see search_messages comment)
         match_stmt = (
             select(models.Message)
             .join(
@@ -868,10 +892,10 @@ async def search_messages_temporal(
         # Order by similarity and limit
         match_stmt = match_stmt.order_by(
             models.MessageEmbedding.embedding.cosine_distance(query_embedding)
-        ).limit(limit)
+        ).limit(limit * 2)
 
         result = await db.execute(match_stmt)
-        matched_messages = list(result.scalars().all())
+        matched_messages = _deduplicate_messages(result.scalars().all(), limit)
     else:
         # External vector store path with post-fetch date filtering
         matched_messages = await _search_messages_external(