fix: six P2 correctness/robustness cleanups from the SDK review

- legacy call_llm: open AsyncOpenAI via `async with` so the client (and
  its HTTP connection pool) is closed instead of leaked.
- LocalBackend.add_document: fail fast with CollectionNotFoundError when
  the collection doesn't exist, before the expensive parse + LLM index
  (previously the missing FK only tripped at save time, after paying for
  the LLM work). Also raise builtin FileNotFoundError for a missing path
  instead of FileTypeError (which now means only "unsupported extension").
- Collection.query(doc_ids=None): the empty-collection guard now always
  runs — previously it was skipped once PAGEINDEX_EXPERIMENTAL_MULTIDOC
  was set. A single list_documents call serves both the guard and the
  multi-doc warning (no separate call just to decide whether to warn).
- CloudBackend.query_stream: on early consumer break / GeneratorExit,
  signal the background SSE thread to stop and force-close the response,
  so it no longer drains the whole stream in the background.

Adds regression tests for each (client closed, fail-fast on unknown
collection, FileNotFoundError, empty-check under the multidoc env flag,
single list call, early-break thread stop).

Claude-Session: https://claude.ai/code/session_01Kx5DgKbhK1N8autqXH8SmS

tests/test_legacy_utils_contract.py

@@ -75,6 +75,7 @@ def test_print_tree_keeps_legacy_exclude_fields(capsys):
 
 def test_call_llm_keeps_legacy_async_openai_contract(monkeypatch):
     calls = []
+    closed = []
 
     class FakeCompletions:
         async def create(self, **kwargs):
@@ -88,6 +89,15 @@ def test_call_llm_keeps_legacy_async_openai_contract(monkeypatch):
             self.api_key = api_key
             self.chat = SimpleNamespace(completions=FakeCompletions())
 
+        # call_llm must open the client as an async context manager so it is
+        # closed (no leaked HTTP connection pool).
+        async def __aenter__(self):
+            return self
+
+        async def __aexit__(self, *exc):
+            closed.append(True)
+            return False
+
     fake_openai = SimpleNamespace(AsyncOpenAI=FakeAsyncOpenAI)
     monkeypatch.setitem(sys.modules, "openai", fake_openai)
 
@@ -99,6 +109,7 @@ def test_call_llm_keeps_legacy_async_openai_contract(monkeypatch):
     ))
 
     assert result == "answer"
+    assert closed == [True]  # client was closed
     assert calls == [{
         "model": "gpt-test",
         "messages": [{"role": "user", "content": "hello"}],