fix(web): VM uptime pinned near 0s while a build was running

uptime_sec came from the clone directory's own mtime, but a running VM's
hypervisor can rewrite lock files under that same directory at any point,
resetting the mtime and pinning "up" near 0s for the VM's entire life --
worst exactly while a build is actively running, since that's when the VM
is busiest. The clone name already carries its own creation timestamp
(_make_clone_name, parsed back via parse_clone_name), set once and never
disturbed after -- prefer it, falling back to the old mtime read only for
a name that doesn't parse.

tests/python/test_operations_vms.py

@@ -321,6 +321,31 @@ def test_list_live_vms_outside_build_vms_not_orphan(tmp_path: Path) -> None:
     assert vm.uptime_sec is None
 
 
+def test_list_live_vms_uptime_uses_clone_name_timestamp_not_directory_mtime(
+    tmp_path: Path,
+) -> None:
+    """The clone directory's own mtime can be disturbed by the hypervisor
+    while the VM is running (VMware rewrites lock files under it) -- the
+    timestamp already baked into the clone's own name at creation time can't
+    be, so it must win over the filesystem mtime."""
+    build_vms = tmp_path / "build-vms"
+    build_vms.mkdir()
+    two_hours_ago = datetime.now(UTC) - timedelta(hours=2)
+    name = f"Clone_412-1_{two_hours_ago:%Y%m%d_%H%M%S}_linux"
+    clone_dir, vmx = _make_clone_dir(build_vms, name)
+    # Something touched the directory moments ago -- its mtime says "just
+    # now" even though the clone itself is 2h old.
+    _age_dir(clone_dir, hours=0)
+
+    backend = _FakeLiveBackend([VmHandle(identifier=str(vmx), name=name)])
+    config = _make_config(build_vms)
+
+    result = list_live_vms(backend, config)  # type: ignore[arg-type]
+    assert len(result) == 1
+    vm = result[0]
+    assert vm.uptime_sec is not None and vm.uptime_sec >= 7000
+
+
 def test_list_live_vms_get_ip_none(tmp_path: Path) -> None:
     build_vms = tmp_path / "build-vms"
     build_vms.mkdir()