refactor(memory): remove flush_memories entirely (#15696)

The AIAgent.flush_memories pre-compression save, the gateway
_flush_memories_for_session, and everything feeding them are
obsolete now that the background memory/skill review handles
persistent memory extraction.

Problems with flush_memories:

- Pre-dates the background review loop.  It was the only memory-save
  path when introduced; the background review now fires every 10 user
  turns on CLI and gateway alike, which is far more frequent than
  compression or session reset ever triggered flush.
- Blocking and synchronous.  Pre-compression flush ran on the live agent
  before compression, blocking the user-visible response.
- Cache-breaking.  Flush built a temporary conversation prefix
  (system prompt + memory-only tool list) that diverged from the live
  conversation's cached prefix, invalidating prompt caching.  The
  gateway variant spawned a fresh AIAgent with its own clean prompt
  for each finalized session — still cache-breaking, just in a
  different process.
- Redundant.  Background review runs in the live conversation's
  session context, gets the same content, writes to the same memory
  store, and doesn't break the cache.  Everything flush_memories
  claimed to preserve is already covered.

What this removes:

- AIAgent.flush_memories() method (~248 LOC in run_agent.py)
- Pre-compression flush call in _compress_context
- flush_memories call sites in cli.py (/new + exit)
- GatewayRunner._flush_memories_for_session + _async_flush_memories
  (and the 3 call sites: session expiry watcher, /new, /resume)
- 'flush_memories' entry from DEFAULT_CONFIG auxiliary tasks,
  hermes tools UI task list, auxiliary_client docstrings
- _memory_flush_min_turns config + init
- #15631's headroom-deduction math in
  _check_compression_model_feasibility (headroom was only needed
  because flush dragged the full main-agent system prompt along;
  the compression summariser sends a single user-role prompt so
  new_threshold = aux_context is safe again)
- The dedicated test files and assertions that exercised
  flush-specific paths

What this renames (with read-time backcompat on sessions.json):

- SessionEntry.memory_flushed -> SessionEntry.expiry_finalized.
  The session-expiry watcher still uses the flag to avoid re-running
  finalize/eviction on the same expired session; the new name
  reflects what it now actually gates.  from_dict() reads
  'expiry_finalized' first, falls back to the legacy 'memory_flushed'
  key so existing sessions.json files upgrade seamlessly.

Supersedes #15631 and #15638.

Tested: 383 targeted tests pass across run_agent/, agent/, cli/,
and gateway/ session-boundary suites.  No behavior regressions —
background memory review continues to handle persistent memory
extraction on both CLI and gateway.

tests/gateway/test_session_boundary_hooks.py

@@ -177,8 +177,8 @@ async def test_idle_expiry_fires_finalize_hook(mock_invoke_hook):
     its reset policy (idle timeout, scheduled reset), it must fire
     ``on_session_finalize`` so plugin providers get the same final-pass
     extraction opportunity they'd get from /new or CLI shutdown.  Before
-    the fix, the expiry path flushed memories and evicted the agent but
-    silently skipped the hook.
+    the fix, the expiry path evicted the agent but silently skipped the
+    hook.
     """
     from datetime import datetime, timedelta
 
@@ -200,7 +200,7 @@ async def test_idle_expiry_fires_finalize_hook(mock_invoke_hook):
         platform=Platform.TELEGRAM,
         chat_type="dm",
     )
-    expired_entry.memory_flushed = False
+    expired_entry.expiry_finalized = False
 
     runner.session_store = MagicMock()
     runner.session_store._ensure_loaded = MagicMock()
@@ -211,24 +211,24 @@ async def test_idle_expiry_fires_finalize_hook(mock_invoke_hook):
     runner.session_store._lock.__exit__ = MagicMock(return_value=None)
     runner.session_store._save = MagicMock()
 
-    runner._async_flush_memories = AsyncMock()
     runner._evict_cached_agent = MagicMock()
     runner._cleanup_agent_resources = MagicMock()
     runner._sweep_idle_cached_agents = MagicMock(return_value=0)
 
     # The watcher starts with `await asyncio.sleep(60)` and loops while
-    # `self._running`. Patch sleep so the 60s initial delay is instant, then
-    # flip `_running` false inside the flush call so the loop exits cleanly
-    # after one pass.
+    # `self._running`.  Patch sleep so the 60s initial delay is instant, and
+    # make the expiry hook invocation flip `_running` false so the loop
+    # exits cleanly after one pass.
     _orig_sleep = __import__("asyncio").sleep
 
     async def _fast_sleep(_):
         await _orig_sleep(0)
 
-    async def _flush_and_stop(session_id, key):
-        runner._running = False  # terminate the loop after this iteration
+    def _hook_and_stop(*a, **kw):
+        runner._running = False
+        return None
 
-    runner._async_flush_memories = AsyncMock(side_effect=_flush_and_stop)
+    mock_invoke_hook.side_effect = _hook_and_stop
 
     with patch("gateway.run.asyncio.sleep", side_effect=_fast_sleep):
         await runner._session_expiry_watcher(interval=0)