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.

docsgpt/agents/tools/internal_search.py

@@ -36,7 +36,7 @@ class InternalSearchTool(Tool):
                 source=self.config.get("source", {}),
                 chat_history=[],
                 prompt="",
-                chunks=int(self.config.get("chunks", 2)),
+                chunks=int(self.config.get("chunks", 6)),
                 doc_token_limit=int(self.config.get("doc_token_limit", 50000)),
                 model_id=self.config.get("model_id", "docsgpt-local"),
                 model_user_id=self.config.get("model_user_id"),
@@ -464,7 +464,7 @@ def add_internal_search_tool(tools_dict: Dict, retriever_config: Dict) -> None:
 def build_internal_tool_config(
     source: Dict,
     retriever_name: str = "classic",
-    chunks: int = 2,
+    chunks: int = 6,
     doc_token_limit: int = 50000,
     sources: Optional[List[Dict]] = None,
     model_id: str = "docsgpt-local",