Default retrieval to 6 chunks instead of 2

chunks is a total per request, split across the attached sources, so an
agent with two sources and the default of 2 got a single chunk from each,
and its answers changed with whichever chunk won. 6 gives three per
source for about 4-5k more input tokens per retrieval turn.

Every literal default moves from 2 to 6: the request default, the
retrievers, the internal search tool, workflow agent nodes, scheduled and
headless runs, agent create/update/import, the source retrieval config and
the frontend forms. Existing agents and sources keep what they store; a
source saved with chunks=2 now counts as configured at 2, which is pinned
by a test.

Headless runs also read chunks=0 as unset (`or 2`), so an agent with
retrieval switched off retrieved anyway on scheduled runs; 0 now stays 0.

tests/api/answer/services/test_stream_processor_utils.py

@@ -473,7 +473,7 @@ class TestConfigureRetriever:
         sp = StreamProcessor({}, {"sub": "u"})
         sp._configure_retriever()
         assert sp.retriever_config["retriever_name"] == "classic"
-        assert sp.retriever_config["chunks"] == 2
+        assert sp.retriever_config["chunks"] == 6
 
     def test_agent_overrides(self):
         from docsgpt.api.answer.services.stream_processor import (
@@ -519,7 +519,7 @@ class TestConfigureRetriever:
         sp._agent_data = {}
         sp._configure_retriever()
         assert sp.retriever_config["retriever_name"] == "classic"
-        assert sp.retriever_config["chunks"] == 2
+        assert sp.retriever_config["chunks"] == 6
 
     def test_invalid_agent_chunks_falls_back(self):
         from docsgpt.api.answer.services.stream_processor import (
@@ -528,7 +528,7 @@ class TestConfigureRetriever:
         sp = StreamProcessor({}, {"sub": "u"})
         sp._agent_data = {"chunks": "not-a-number"}
         sp._configure_retriever()
-        assert sp.retriever_config["chunks"] == 2
+        assert sp.retriever_config["chunks"] == 6
 
     def test_invalid_request_chunks_falls_back(self):
         from docsgpt.api.answer.services.stream_processor import (
@@ -536,7 +536,7 @@ class TestConfigureRetriever:
         )
         sp = StreamProcessor({"chunks": "abc"}, {"sub": "u"})
         sp._configure_retriever()
-        assert sp.retriever_config["chunks"] == 2
+        assert sp.retriever_config["chunks"] == 6
 
     def test_isnonedoc_without_api_key_sets_chunks_to_0(self):
         from docsgpt.api.answer.services.stream_processor import (

tests/api/answer/test_retrieved_docs_reach_the_model.py

@@ -126,6 +126,15 @@ class TestChunksPrecedence:
         sp._configure_retriever()
         return sp.retriever_config["chunks"]
 
+    def test_default_applies_when_nothing_sets_chunks(self):
+        assert self._sp() == 6
+
+    def test_source_stored_at_the_old_default_counts_as_configured(self):
+        """Accepted when the default moved from 2 to 6: sources created while
+        the UI always sent the full config store ``chunks=2`` explicitly, and
+        stay a per-source setting of 2 rather than being migrated."""
+        assert self._sp(request_chunks="7", source_chunks=2) == 2
+
     def test_request_applies_when_source_is_unconfigured(self):
         assert self._sp(request_chunks="7") == 7
 
@@ -139,7 +148,7 @@ class TestChunksPrecedence:
             ("-5", 0),
             ("100000", 500),
             ("501", 500),
-            ("abc", 2),
+            ("abc", 6),
         ],
     )
     def test_request_chunks_is_clamped(self, sent, expected):

tests/api/answer/test_stream_processor.py

@@ -1202,7 +1202,7 @@ class TestConfigureAgent:
         assert sp.retriever_config["chunks"] == 5
 
     @pytest.mark.unit
-    def test_configure_agent_invalid_chunks_defaults_to_2(self):
+    def test_configure_agent_invalid_chunks_uses_the_default(self):
         sp = self._make_sp()
         sp._resolve_agent_id = MagicMock(return_value="agent_id_1")
         sp._get_agent_key = MagicMock(return_value=("agent_key", False, None))
@@ -1220,7 +1220,7 @@ class TestConfigureAgent:
         sp._configure_agent()
         sp.model_id = "test-model"
         sp._configure_retriever()
-        assert sp.retriever_config["chunks"] == 2
+        assert sp.retriever_config["chunks"] == 6
 
 
 # ---- Additional coverage: _load_conversation_history ----

tests/api/test_agent_portability.py

@@ -176,6 +176,20 @@ def test_import_by_slug_idempotent(pg_conn):
     assert len(AgentsRepository(pg_conn).list_for_user(user)) == 1
 
 
+def test_import_without_chunks_uses_the_default(pg_conn):
+    user = "u_default_chunks"
+    result = apply_import(pg_conn, user, _doc(name="Plain Bot", _slug="plain-bot"))
+    agent = AgentsRepository(pg_conn).get(result["agent_id"], user)
+    assert agent["chunks"] == 6
+
+
+def test_import_keeps_explicit_chunks(pg_conn):
+    user = "u_explicit_chunks"
+    result = apply_import(pg_conn, user, _doc(name="Off Bot", _slug="off-bot", chunks=0))
+    agent = AgentsRepository(pg_conn).get(result["agent_id"], user)
+    assert agent["chunks"] == 0
+
+
 def test_import_missing_source_drafts_and_warns(pg_conn):
     user = "u_missing"
     doc = _doc(

tests/api/user/agents/test_routes_happy.py

@@ -426,6 +426,31 @@ class TestCreateAgent:
         agents = AgentsRepository(pg_conn).list_for_user(user)
         assert any(a["name"] == "My Draft Agent" for a in agents)
 
+    def test_agent_created_without_chunks_stores_the_default(self, app, pg_conn):
+        # With two sources, 2 left each one a single chunk, so the answer hung
+        # on which chunk won; 6 is the default now.
+        from docsgpt.api.user.agents.routes import CreateAgent
+        from docsgpt.storage.db.repositories.agents import AgentsRepository
+
+        user = "u-create-default-chunks"
+
+        with _patch_db(pg_conn), app.test_request_context(
+            "/api/create_agent",
+            method="POST",
+            json={
+                "name": "Handbook Bot",
+                "description": "d",
+                "agent_type": "classic",
+                "status": "draft",
+            },
+        ):
+            from flask import request
+            request.decoded_token = {"sub": user}
+            response = CreateAgent().post()
+        assert response.status_code == 201
+        (agent,) = AgentsRepository(pg_conn).list_for_user(user)
+        assert agent["chunks"] == 6
+
 
 # ---------------------------------------------------------------------------
 # UpdateAgent — big method with many validation branches