## What happened

After an idle cycle whose first "active" signal comes from the screensaver
launch itself, `omarchy-shell idle status` keeps reporting `inIdleCycle: true`
for the rest of the session, including after the user returns and unlocks.
Anything that reads `inIdleCycle` as "the user is idle" then acts on a stale
flag. In my case a plugin suspended the machine every few minutes while I was
using it, and only a reboot stopped it.

## Cause

In `plugins/services/idle/Service.qml`, `handleActiveSignal()` ignores an
"active" signal while the screensaver cycle is armed:

    if (root.screensaverStartedThisCycle && (root.screensaverWindowCount > 0 || screensaverLaunchGraceTimer.running)) {
      logEvent("idle-monitor-active", "screensaver cycle remains armed")
      return
    }

The launch-time "active" falls inside the grace window, so it is ignored. The
monitor is now in the active state and emits no further transition when the
user really returns, so `cancelIdleCycle("activity")` never runs. No
screensaver window ever appears, so the window-closed cancel path doesn't run
either. `idledThisCycle` stays true.

## Log (journalctl --user, idle debug logging)

    07:08:11  idle-monitor: idle
    07:08:11  idle-cycle-start: screensaver=3600 lock=7200
    07:08:12  idle-monitor: active
    07:08:12  idle-monitor-active: screensaver cycle remains armed
    07:08:12  process-exit: screensaver exitCode=0 status=0
    08:08:11  lock-system: lock-timeout
    ...
    10:21:45  lock: secure=false / unlocked
    (no idle-cycle-cancel and no later idle-monitor event for the rest of the boot)

Nine seconds after the 10:21:45 unlock, a consumer still saw the cycle as
active and ran `systemctl suspend`.

## Expected

The cycle is cancelled once the user is active again, or at least cleared on
unlock. `inIdleCycle` should not outlive the user's return.

## Notes

- Related but distinct: #12225 (screensaver self-exits and cancels the cycle
  too early). This is the opposite failure: the cycle is never cancelled.
- Consumers can defend against it by also requiring `idle: true` or
  `screensaverWindows > 0`.

Omarchy 4.0.4-1, kernel 7.2.5-3-omarchy, Hyprland 0.56.2, quickshell 0.3.2,
4 monitors, idle: {screensaver: 3600, lock: 7200}

Filed by Claude Sonnet 5.5 via Claude Code