fix(backends): make streamed replies and session state consistent

- Validate stage prefixes before forwarding model tokens. Invalid prefixes
  may use the existing repair budget; once output starts, validation failure
  aborts the turn without appending a regenerated reply. Bound the token
  queue and close inner streams when consumption stops.
- Replace delete-time lock eviction with reference-counted session locks
  covering holders and waiters. Serialize set/get with chat/stream/delete
  and release cancelled waiters without replacing a live lock.
- Store person synonyms under workflow-native driver keys and expose
  consistent alias views to single/group reads and graph prompt injection.
- Separate committed field patches from public formUpdate visibility.
  Commit callbacks keep repository fields synchronized even when updates
  are hidden or the consumer closes after graph completion.
- Disable OpenAI SDK automatic retries. Own bounded retry attempts in the
  application, use 1/2/4 second backoff, retry streams only before text is
  emitted, enforce attempt timeouts, and close upstream/client resources.
- Translate FastGPT SDK max_retries (total attempts) from the configured
  extra-retry count, and verify the real SDK loop uses four attempts.
- Document the per-LLM budget and streaming failure semantics without
  claiming the whole multi-call HTTP turn has the same timeout.

Validation: 267 offline tests passed, including actual SDK requests through
MockTransport, stream repair/abort, timeout/close, lock races, aliases and
formUpdate visibility; existing four dependency deprecation warnings.

test/backends/test_langgraph_streaming.py

@@ -58,7 +58,7 @@ class StreamingFakeLLM:
 
 
 def make_input(session="stream-s1", text="继续办理"):
-    return ChatInput(session_id=session, text=text, client_mode="direct")
+    return ChatInput(session_id=session, text=text, client_mode="direct", need_form_update=True)
 
 
 async def _second_turn_events(backend, session="stream-s1"):