fix(monitor): Stop re-promoting a stale notification into a new session

app/src/main/java/eu/darken/capod/monitor/core/worker/MonitorService.kt

@@ -184,7 +184,15 @@ class MonitorService : Service() {
     override fun onStartCommand(intent: Intent?, flags: Int, startId: Int): Int {
         // Every startForegroundService() re-arms the startForeground() obligation,
         // so every onStartCommand has to satisfy it again, no matter how it exits.
-        if (!promoteToForeground(lastNotification ?: MonitorNotifications.createEarlyNotification(this))) {
+        //
+        // Only reuse the cached notification while a monitor session is actually live. It is a fully
+        // built object whose `when` is frozen at build time, so re-promoting it to open a NEW session
+        // would re-post the previous session's last content under its original timestamp — which is
+        // how a stale "unknown device" frame survives a teardown/restart cycle. Rebuilding is not an
+        // option here: this has to stay cheap enough to satisfy the obligation before DI is ready.
+        val reusable = lastNotification?.takeIf { monitoringJob?.isActive == true }
+        if (reusable == null) lastNotification = null
+        if (!promoteToForeground(reusable ?: MonitorNotifications.createEarlyNotification(this))) {
             stopSelf(startId)
             return START_NOT_STICKY
         }

app/src/test/java/eu/darken/capod/monitor/core/worker/MonitorServiceTest.kt

@@ -125,6 +125,28 @@ class MonitorServiceTest {
         shadowOf(service).lastForegroundNotificationId shouldBe MonitorNotifications.NOTIFICATION_ID
     }
 
+    /**
+     * A built Notification's `when` is frozen, so re-promoting the previous session's last frame to
+     * open a new one re-posts stale content under its original timestamp — the way an "unknown
+     * device" placeholder survives a teardown/restart cycle.
+     */
+    @Test
+    fun `a start with no live monitor does not reuse the previous notification`() {
+        val service = createService()
+        service.readyForMonitoring()
+
+        val stale = notification("stale")
+        service.postPrimaryNotification(stale)
+
+        // Monitor session has ended — the cached frame no longer reflects anything current.
+        service.setField("monitoringJob", null)
+
+        service.onStartCommand(MonitorService.intent(context), 0, 1) shouldBe Service.START_STICKY
+
+        shadowOf(service).lastForegroundNotification shouldNotBeSameInstanceAs stale
+        service.getField("lastNotification") shouldNotBeSameInstanceAs stale
+    }
+
     /**
      * The FGS notification can outlive `stopSelf()`. Leaving it up strands whatever content was last
      * posted — including the unknown-device placeholder built before the first BLE scan batch landed.