Port from cline/cline#10343: periodic gateway memory logging (#27102)

Emit a grep-friendly '[MEMORY] rss=...MB ...' line in agent.log /
gateway.log every N minutes (default 5) so slow leaks in the long-lived
gateway process show up as a time series. Based on
https://github.com/cline/cline/pull/10343
(src/standalone/memory-monitor.ts).

- gateway/memory_monitor.py: new module. Daemon thread, baseline on
  start, final snapshot on stop. Uses resource.getrusage() (stdlib)
  first, falls back to psutil, disables itself with one WARNING if
  neither is available.
- gateway/run.py: start monitor right after setup_logging() in
  start_gateway(); stop it in the shutdown block next to MCP teardown.
- hermes_cli/config.py: logging.memory_monitor { enabled, interval_seconds }
  defaults under the existing logging section.
- tests/gateway/test_memory_monitor.py: 10 unit tests covering format,
  baseline/shutdown snapshots, double-start noop, periodic timer,
  daemon thread invariant, and unavailable-RSS warn-and-skip path.

Adapted from TypeScript/Node to Python (threading.Event-based daemon
thread instead of setInterval/unref), added Python-specific gc + thread
counts to the log line (handier than ext/arrayBuffers for diagnosing
Python gateway leaks), and gated behind a config.yaml toggle so users
can silence the periodic line if they want.

No heap-snapshot-on-OOM equivalent — CPython doesn't have V8's
--heapsnapshot-near-heap-limit; tracemalloc would be the Python
equivalent but adds non-trivial overhead, so leaving that out.

gateway/run.py

@@ -16800,6 +16800,33 @@ async def start_gateway(config: Optional[GatewayConfig] = None, replace: bool =
     from hermes_logging import setup_logging
     setup_logging(hermes_home=_hermes_home, mode="gateway")
 
+    # Periodic process memory usage logging (gateway only) — emits a
+    # grep-friendly "[MEMORY] rss=...MB ..." line every N minutes so
+    # slow leaks in the long-lived gateway process show up as a time
+    # series in agent.log / gateway.log.  Ported from cline/cline#10343.
+    # Controlled by the logging.memory_monitor section in config.yaml.
+    try:
+        from gateway import memory_monitor as _memory_monitor
+
+        _mm_cfg = {}
+        try:
+            # config is loaded a few lines up; re-read the logging section
+            # here so we pick up user overrides without coupling to local
+            # variable names inside the start_gateway body.
+            from hermes_cli.config import load_config as _load_cli_config
+
+            _mm_cfg = (_load_cli_config() or {}).get("logging", {}).get("memory_monitor", {}) or {}
+        except Exception:
+            _mm_cfg = {}
+        if _mm_cfg.get("enabled", True):
+            try:
+                _mm_interval = float(_mm_cfg.get("interval_seconds", 300))
+            except (TypeError, ValueError):
+                _mm_interval = 300.0
+            _memory_monitor.start_memory_monitoring(interval_seconds=_mm_interval)
+    except Exception as _mm_exc:
+        logger.debug("Failed to start memory monitor: %s", _mm_exc)
+
     # Optional stderr handler — level driven by -v/-q flags on the CLI.
     # verbosity=None (-q/--quiet): no stderr output
     # verbosity=0    (default):    WARNING and above
@@ -17016,6 +17043,16 @@ async def start_gateway(config: Optional[GatewayConfig] = None, replace: bool =
     except Exception:
         pass
 
+    # Stop the periodic memory monitor (if it was started above).
+    # This also emits one final "[MEMORY] shutdown rss=..." line so the
+    # last RSS reading before gateway exit is always in the log.
+    try:
+        from gateway import memory_monitor as _memory_monitor
+
+        _memory_monitor.stop_memory_monitoring(timeout=2.0)
+    except Exception:
+        pass
+
     if runner.exit_code is not None:
         raise SystemExit(runner.exit_code)