fix: cap sqlite-vec knn k parameter at 4096 limit

sqlite-vec enforces k <= 4096 for nearest-neighbor queries. Projects with
>4096 vector chunks would crash all vector/hybrid search because
candidate_limit = max(100, (limit + offset) * 10) exceeded this hard limit.

Clamp the knn k in _run_vector_query while keeping the outer SQL LIMIT
unclamped. Only affects SQLite — pgvector has no such constraint.

Fixes #604

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Signed-off-by: phernandez <paul@basicmachines.co>

src/basic_memory/repository/sqlite_search_repository.py

@@ -438,12 +438,17 @@ class SQLiteSearchRepository(SearchRepositoryBase):
         """Load sqlite-vec extension for the session."""
         await self._ensure_sqlite_vec_loaded(session)
 
+    # sqlite-vec hard limit for knn k parameter
+    SQLITE_VEC_MAX_K = 4096
+
     async def _run_vector_query(
         self,
         session: AsyncSession,
         query_embedding: list[float],
         candidate_limit: int,
     ) -> list[dict]:
+        # Constraint: sqlite-vec enforces k <= 4096 for knn queries
+        vector_k = min(candidate_limit, self.SQLITE_VEC_MAX_K)
         query_embedding_json = json.dumps(query_embedding)
         vector_result = await session.execute(
             text(
@@ -458,12 +463,13 @@ class SQLiteSearchRepository(SearchRepositoryBase):
                 "JOIN search_vector_chunks c ON c.id = vector_matches.rowid "
                 "WHERE c.project_id = :project_id "
                 "ORDER BY best_distance ASC "
-                "LIMIT :vector_k"
+                "LIMIT :candidate_limit"
             ),
             {
                 "query_embedding": query_embedding_json,
                 "project_id": self.project_id,
-                "vector_k": candidate_limit,
+                "vector_k": vector_k,
+                "candidate_limit": candidate_limit,
             },
         )
         return [dict(row) for row in vector_result.mappings().all()]

tests/repository/test_sqlite_vector_search_repository.py

@@ -1,6 +1,8 @@
 """SQLite sqlite-vec search repository tests."""
 
+import json
 from datetime import datetime, timezone
+from unittest.mock import AsyncMock, MagicMock
 
 import pytest
 from sqlalchemy import text
@@ -264,3 +266,52 @@ async def test_sqlite_hybrid_search_combines_fts_and_vector(search_repository):
 
     assert results
     assert any(result.permalink == "specs/search-index" for result in results)
+
+
+@pytest.mark.asyncio
+async def test_run_vector_query_caps_k_at_sqlite_vec_limit(search_repository):
+    """_run_vector_query must cap the knn k param at SQLITE_VEC_MAX_K (4096).
+
+    sqlite-vec raises OperationalError when k > 4096. The candidate_limit
+    passed from the base class can exceed this for large projects, so
+    _run_vector_query clamps k while keeping the outer LIMIT unclamped.
+    """
+    if not isinstance(search_repository, SQLiteSearchRepository):
+        pytest.skip("sqlite-vec k limit is SQLite-specific.")
+
+    _enable_semantic(search_repository)
+    await search_repository.init_search_index()
+
+    # Track the parameters passed to session.execute
+    captured_params: list[dict] = []
+    original_execute = None
+
+    async def capturing_execute(stmt, params=None):
+        if params and "vector_k" in params:
+            captured_params.append(dict(params))
+        # Return empty result set
+        mock_result = MagicMock()
+        mock_result.mappings.return_value.all.return_value = []
+        return mock_result
+
+    async with db.scoped_session(search_repository.session_maker) as session:
+        await search_repository._prepare_vector_session(session)
+        original_execute = session.execute
+        session.execute = capturing_execute
+
+        query_embedding = [0.1] * search_repository._vector_dimensions
+
+        # candidate_limit exceeds sqlite-vec limit
+        await search_repository._run_vector_query(session, query_embedding, 10000)
+
+        assert len(captured_params) == 1
+        assert captured_params[0]["vector_k"] == SQLiteSearchRepository.SQLITE_VEC_MAX_K
+        assert captured_params[0]["candidate_limit"] == 10000
+
+        # candidate_limit within limit should pass through unchanged
+        captured_params.clear()
+        await search_repository._run_vector_query(session, query_embedding, 500)
+
+        assert len(captured_params) == 1
+        assert captured_params[0]["vector_k"] == 500
+        assert captured_params[0]["candidate_limit"] == 500