fix(siteruntime): arm stuck-script watchdog + health/summary on alarm runs (arch-review 2026-08-19 P1-3)

Alarm on-trigger runs (ScriptRunLauncher.RunAlarmScriptAsync) bypassed the
stuck-script watchdog, the script-error health counter, and
ScriptRunSummaryRecorder entirely, even though they run on the same bounded
ScriptExecutionScheduler pool as instance scripts. A wedged alarm body cost a
pool thread forever with no site event and no detach/replace, and a timed-out
or failed on-trigger run was invisible to health monitoring and the run
summary.

Bring the alarm path to parity with the instance-script path (RunScriptAsync):
capture the worker slot/run stamp in the first synchronous segment and arm
ArmStuckScriptWatchdog with the same detach-and-replace semantics and gauge;
resolve ScriptRunSummaryRecorder/ISiteEventLogger from the same DI provider and
thread RecordStarted/RecordCompleted/RecordFailed/RecordTimedOut; increment
ISiteHealthCollector.IncrementScriptError on both the timeout and general
failure catch blocks. LaunchAlarmScript/RunAlarmScriptAsync now take
healthCollector and serviceProvider parameters, passed from AlarmActor's
existing _healthCollector/_serviceProvider fields. No change to the
enqueue-anchored deadline, the shed cap, or "a stop does NOT cancel an
in-flight run" semantics.

Adds AlarmScriptWatchdogTests (previously zero coverage of LaunchAlarmScript):
wedged-worker detach + replacement, timeout-at-dequeue health counter
increment, and successful-completion summary recording. All three fail
against the pre-fix code (verified via stash) and pass after.

src/ZB.MOM.WW.ScadaBridge.SiteRuntime/Actors/AlarmActor.cs

@@ -825,6 +825,11 @@ public class AlarmActor : ReceiveActor
                 // Context.Parent.Tell from the execution actor.
                 Self,
                 _logger,
+                // arch-review 2026-08-19 P1-3: same health collector + service provider as
+                // the instance-script path, so the alarm on-trigger run gets the same
+                // stuck-script watchdog, script-error counting, and run-summary threading.
+                _healthCollector,
+                _serviceProvider,
                 runId,
                 // Per-script timeout from the on-trigger script (null = global).
                 _onTriggerExecutionTimeoutSeconds,

src/ZB.MOM.WW.ScadaBridge.SiteRuntime/Scripts/ScriptRunLauncher.cs

@@ -138,6 +138,18 @@ internal static class ScriptRunLauncher
     /// <param name="options">Site runtime configuration options, including the execution timeout.</param>
     /// <param name="completionTarget">The owning <see cref="AlarmActor"/>, which receives the completion message.</param>
     /// <param name="logger">Logger for execution diagnostics.</param>
+    /// <param name="healthCollector">
+    /// Optional health collector for recording execution metrics (arch-review 2026-08-19
+    /// P1-3): the alarm on-trigger run shares the script-error counter with
+    /// <see cref="RunScriptAsync"/> so a wedged/failed alarm body is visible the same way a
+    /// wedged/failed instance script is.
+    /// </param>
+    /// <param name="serviceProvider">
+    /// Optional DI service provider (arch-review 2026-08-19 P1-3): resolves the singleton
+    /// <see cref="ISiteEventLogger"/> (stuck-watchdog reporting) and
+    /// <see cref="ScriptRunSummaryRecorder"/> (aggregate run counters), exactly like the
+    /// instance-script path.
+    /// </param>
     /// <param name="runId">Per-alarm run counter, used only in log messages.</param>
     /// <param name="executionTimeoutSeconds">The on-trigger script's per-script timeout in seconds. Null or non-positive falls back to the global value.</param>
     /// <param name="parentExecutionId">
@@ -158,6 +170,8 @@ internal static class ScriptRunLauncher
         SiteRuntimeOptions options,
         IActorRef completionTarget,
         ILogger logger,
+        ISiteHealthCollector? healthCollector,
+        IServiceProvider? serviceProvider,
         long runId,
         int? executionTimeoutSeconds,
         Guid? parentExecutionId,
@@ -174,7 +188,8 @@ internal static class ScriptRunLauncher
                 () => RunAlarmScriptAsync(
                     alarmName, instanceName, level, priority, message, compiledScript,
                     instanceActor, sharedScriptLibrary, options, completionTarget, logger,
-                    runId, parentExecutionId, timeout, cts),
+                    healthCollector, serviceProvider, runId, parentExecutionId, timeout,
+                    executionScheduler, cts),
                 CancellationToken.None, TaskCreationOptions.DenyChildAttach, executionScheduler)
                 .Unwrap();
         }
@@ -463,11 +478,42 @@ internal static class ScriptRunLauncher
         SiteRuntimeOptions options,
         IActorRef completionTarget,
         ILogger logger,
+        ISiteHealthCollector? healthCollector,
+        IServiceProvider? serviceProvider,
         long runId,
         Guid? parentExecutionId,
         TimeSpan timeout,
+        ScriptExecutionScheduler executionScheduler,
         CancellationTokenSource cts)
     {
+        // arch-review 2026-08-19 P1-3: singletons resolved from the root provider up front,
+        // exactly like RunScriptAsync, so they are available to the catch blocks regardless
+        // of scope state. Null in tests that pass no serviceProvider — degrades the same way
+        // the instance-script path does (no site event log / no aggregate counters).
+        var siteEventLogger = serviceProvider?.GetService<ISiteEventLogger>();
+        var summaryRecorder = serviceProvider?.GetService<ScriptRunSummaryRecorder>();
+
+        // The on-trigger run has no per-script name of its own in the site-event/summary
+        // sense — key it under the alarm, mirroring the "AlarmActor:{name}" source
+        // convention already used for this alarm's other site events.
+        var scriptName = $"AlarmActor:{alarmName}";
+        var runStopwatch = Stopwatch.StartNew();
+
+        // arch-review 2026-08-19 P1-3: identify the worker this body is holding, so the
+        // stuck-script watchdog can detach and replace it if the body wedges — same
+        // mechanism, same gauge, as the instance-script path. Read in the FIRST synchronous
+        // segment, before any await.
+        var workerSlot = ScriptExecutionScheduler.CurrentWorkerSlot;
+        var workerRunStamp = workerSlot is { } slot ? executionScheduler.CurrentRunStamp(slot) : 0L;
+
+        var completed = 0;
+        ArmStuckScriptWatchdog(
+            cts, options, logger, siteEventLogger, executionScheduler,
+            workerSlot, workerRunStamp, timeout,
+            () => Volatile.Read(ref completed) != 0,
+            $"Alarm on-trigger script for '{alarmName}' on instance '{instanceName}'",
+            instanceName, scriptName);
+
         try
         {
             // Enqueue-anchored deadline: an already-cancelled token means this on-trigger run
@@ -511,12 +557,17 @@ internal static class ScriptRunLauncher
                 }
             };
 
+            summaryRecorder?.RecordStarted(instanceName, scriptName);
+
             await compiledScript.RunAsync(globals, cts.Token);
 
+            summaryRecorder?.RecordCompleted(instanceName, scriptName, runStopwatch.ElapsedMilliseconds);
             completionTarget.Tell(new AlarmActor.AlarmExecutionCompleted(alarmName, true));
         }
         catch (OperationCanceledException)
         {
+            healthCollector?.IncrementScriptError();
+            summaryRecorder?.RecordTimedOut(instanceName, scriptName, runStopwatch.ElapsedMilliseconds);
             logger.LogWarning(
                 "Alarm on-trigger script for {Alarm} on {Instance} timed out (run {RunId})",
                 alarmName, instanceName, runId);
@@ -524,6 +575,8 @@ internal static class ScriptRunLauncher
         }
         catch (Exception ex)
         {
+            healthCollector?.IncrementScriptError();
+            summaryRecorder?.RecordFailed(instanceName, scriptName, runStopwatch.ElapsedMilliseconds);
             // Failures logged, alarm continues
             logger.LogError(ex,
                 "Alarm on-trigger script for {Alarm} on {Instance} failed (run {RunId})",
@@ -532,6 +585,9 @@ internal static class ScriptRunLauncher
         }
         finally
         {
+            // Mark the body finished so the stuck-script watchdog (registered on
+            // cts.Token above) treats a timely cancellation as NOT stuck.
+            Interlocked.Exchange(ref completed, 1);
             cts.Dispose();
         }
     }