fix(gateway): use process-level HERMES_HOME for identity files (#56993 salvage) (#59341)

* fix(gateway): use process-level HERMES_HOME for identity files

Gateway identity files (PID, lock, runtime status, takeover/stop markers)
were written via get_hermes_home() which honours the _HERMES_HOME_OVERRIDE
contextvar used for per-session profile dispatch.  When a profile-context
task happened to be active at write time, files landed in the wrong profile
directory.

Add _get_process_hermes_home() that skips the contextvar and uses only the
HERMES_HOME env var or platform default, and route all gateway identity file
paths through it.

Fixes #56986

* chore(release): map liuhao1024 author email for PR #56993 salvage

---------

Co-authored-by: liuhao1024 <sunsky.lau@gmail.com>
Co-authored-by: Ben <ben@nousresearch.com>

gateway/status.py

@@ -20,7 +20,7 @@ import subprocess
 import sys
 from datetime import datetime, timezone
 from pathlib import Path
-from hermes_constants import get_hermes_home
+from hermes_constants import get_hermes_home, _get_platform_default_hermes_home
 from typing import Any, Optional
 from utils import atomic_json_write
 
@@ -42,9 +42,25 @@ _gateway_lock_handle = None
 _WINDOWS_LOCK_OFFSET = 1024 * 1024
 
 
+def _get_process_hermes_home() -> Path:
+    """Return the process-level HERMES_HOME, skipping context-local overrides.
+
+    Gateway identity files (PID, lock, runtime status, takeover/stop markers)
+    must always live in the directory the gateway process was launched with.
+    ``get_hermes_home()`` honors ``_HERMES_HOME_OVERRIDE`` contextvar used for
+    per-session profile dispatch, which would route these files into the wrong
+    profile directory when a profile-context task happens to be active at write
+    time.  See issue #56986.
+    """
+    val = os.environ.get("HERMES_HOME", "").strip()
+    if val:
+        return Path(val)
+    return _get_platform_default_hermes_home()
+
+
 def _get_pid_path() -> Path:
     """Return the path to the gateway PID file, respecting HERMES_HOME."""
-    home = get_hermes_home()
+    home = _get_process_hermes_home()
     return home / "gateway.pid"
 
 
@@ -52,7 +68,7 @@ def _get_gateway_lock_path(pid_path: Optional[Path] = None) -> Path:
     """Return the path to the runtime gateway lock file."""
     if pid_path is not None:
         return pid_path.with_name(_GATEWAY_LOCK_FILENAME)
-    home = get_hermes_home()
+    home = _get_process_hermes_home()
     return home / _GATEWAY_LOCK_FILENAME
 
 
@@ -1136,13 +1152,13 @@ _PLANNED_STOP_MARKER_TTL_S = 60
 
 def _get_takeover_marker_path() -> Path:
     """Return the path to the --replace takeover marker file."""
-    home = get_hermes_home()
+    home = _get_process_hermes_home()
     return home / _TAKEOVER_MARKER_FILENAME
 
 
 def _get_planned_stop_marker_path() -> Path:
     """Return the path to the intentional gateway stop marker file."""
-    home = get_hermes_home()
+    home = _get_process_hermes_home()
     return home / _PLANNED_STOP_MARKER_FILENAME
 
 
@@ -1197,7 +1213,7 @@ def _consume_pid_marker_for_self(
     # unaffected. Leave a mismatched marker in place so the correct
     # profile can still consume it.
     replacer_home = record.get("replacer_hermes_home")
-    if replacer_home is not None and replacer_home != str(get_hermes_home()):
+    if replacer_home is not None and replacer_home != str(_get_process_hermes_home()):
         return False
 
     our_pid = os.getpid()
@@ -1246,7 +1262,7 @@ def write_takeover_marker(target_pid: int) -> bool:
             "target_pid": target_pid,
             "target_start_time": target_start_time,
             "replacer_pid": os.getpid(),
-            "replacer_hermes_home": str(get_hermes_home()),
+            "replacer_hermes_home": str(_get_process_hermes_home()),
             "written_at": _utc_now_iso(),
         }
         _write_json_file(_get_takeover_marker_path(), record)

scripts/release.py

@@ -53,6 +53,7 @@ AUTHOR_MAP = {
     "allenliang2022@users.noreply.github.com": "allenliang2022",  # PR #56932 test coverage folded into #56909 salvage (408 → retryable timeout)
     "m888.braun@hotmail.com": "ManniBr",  # PR #57417 partial salvage (gateway: fail-closed adapter resolution for unregistered secondary profiles)
     "austin@openvm067.space": "austinlaw076",  # PR #57563 partial salvage (auth: lazy per-profile Anthropic OAuth file; gateway: whatsapp_cloud/line added to port-binding platform set)
+    "sunsky.lau@gmail.com": "liuhao1024",  # PR #56993 salvage (gateway: process-level HERMES_HOME for pid/lock/status identity files; #56986)
     "blueirobin02@gmail.com": "irresi",  # PR #59048 salvage (gateway: scope reset banners' session info to the serving profile; #59003)
     "jashlee+microsoft@microsoft.com": "s905060",  # PR #57943 salvage (photon: auto-reinstall stale sidecar node_modules when lockfile is newer than npm's install marker; #59169)
     "lohinth25@proton.me": "l0h1nth",  # PR #32210 salvage (mattermost: accept leading-space slash commands from mobile clients; #25184)

tests/gateway/test_status.py

@@ -293,6 +293,41 @@ class TestGatewayPidState:
         finally:
             status.release_gateway_runtime_lock()
 
+    def test_gateway_identity_files_use_process_home_not_context_override(
+        self, tmp_path, monkeypatch
+    ):
+        """Regression: pid/lock/state files must use process-level HERMES_HOME.
+
+        When a profile context override is active (e.g., during session dispatch
+        for a named profile), gateway identity files should still be written to
+        the process-level HERMES_HOME, not the profile's directory.  See #56986.
+        """
+        from hermes_constants import set_hermes_home_override, reset_hermes_home_override
+
+        process_home = tmp_path / "default"
+        process_home.mkdir()
+        profile_home = tmp_path / "profiles" / "cfo"
+        profile_home.mkdir(parents=True)
+        monkeypatch.setenv("HERMES_HOME", str(process_home))
+
+        # Simulate a profile context override being active during write.
+        token = set_hermes_home_override(str(profile_home))
+        try:
+            status.write_pid_file()
+        finally:
+            reset_hermes_home_override(token)
+
+        # PID file must land in the process-level home, not the profile home.
+        assert (process_home / "gateway.pid").exists()
+        assert not (profile_home / "gateway.pid").exists()
+
+        payload = json.loads((process_home / "gateway.pid").read_text())
+        assert payload["pid"] == os.getpid()
+
+        # Cleanup for atexit hooks.
+        monkeypatch.setenv("HERMES_HOME", str(process_home))
+        (process_home / "gateway.pid").unlink(missing_ok=True)
+
 
 class TestGatewayRuntimeStatus:
     def test_write_json_file_uses_atomic_json_write(self, tmp_path, monkeypatch):