Resolve all 33 medium-severity architecture review findings

Second slice of the 2026-09-09 whole-system review remediation
(docs/plans/2026-09-09-arch-review.md §10), on top of the High slice 146a1323.

Runtime/devices: SR-2 alarm on-trigger integration scope; SR-3 correlated NativeAlarmActor
subscribe retry; SR-4 immutable audited DB target; SR-5 ordered InstanceActor stop/flush;
SR-12 rehydration fence; SR-13 S&F gate recheck per delivery; DC-4 no OPC UA security
downgrade; DC-5 alarm snapshot for every new subscriber; ORCH-1 sync-dispose joins the pump.

Management/UI: MG-2 site scope on audit tree + Flow Debugger Replay; MG-3 stale-probe memo
removed; MG-4 server-owned bulk-deploy deadline; UI-5 sandbox runs in a killable child
process (new CentralUI.SandboxWorker project); UI-14 bounded sandbox output/IPC.

Edge/audit: EI-1 relay audit ownership; EI-2 SkipBodyCapture suppresses error text; EI-3/EI-11
error-field redaction + caps; EI-4 single relay deadline, pre-execution-only retry; EI-5
authenticated relay provenance; EI-10 body cap before audit read.

Persistence/contracts: DL-1 global Instance.UniqueName unique index + conflict translation;
DL-2 DataConnection protocol config encrypted at rest (migration
20260909203046_EnforceGlobalInstanceNamesAndProtectProtocolConfiguration — coordinated
central upgrade, old binaries cannot read protected rows); DL-3 bounded Replay paging;
DL-6 cross-site Overwrite of a deployed instance rejected; XC-4 proto descriptor contract test.

Config/hygiene: XC-1 central listener isolation by local port; XC-2/XC-3 inert SMTP + cron
keys removed from rig settings; CM-2 env2 bootstrap guard enabled; DC-6 deploy site exit code;
DC-7 notifier retries only explicit standby rejection; XC-5 ClusterClient purged from docs.

Gates recorded in the report: build 0/0, 9 397 passed / 0 failed, rig redeployed on all
eight nodes with the migration applied. 46 Low findings remain open.

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

@@ -165,6 +165,7 @@ internal static class ScriptRunLauncher
     /// (external data has no spawning execution) — that on-trigger run is a tree ROOT.
     /// </param>
     /// <param name="scheduler">Script-execution scheduler seam (#18); null selects the process-wide shared instance.</param>
+    /// <param name="siteCommunicationActor">Site transport for notification status queries, resolved on the actor thread.</param>
     public static void LaunchAlarmScript(
         string alarmName,
         string instanceName,
@@ -182,7 +183,8 @@ internal static class ScriptRunLauncher
         long runId,
         int? executionTimeoutSeconds,
         Guid? parentExecutionId,
-        ScriptExecutionScheduler? scheduler)
+        ScriptExecutionScheduler? scheduler,
+        ICanTell? siteCommunicationActor = null)
     {
         var timeout = ResolveTimeout(executionTimeoutSeconds, options);
         var executionScheduler = scheduler ?? ScriptExecutionScheduler.Shared(options);
@@ -196,7 +198,7 @@ internal static class ScriptRunLauncher
                     alarmName, instanceName, level, priority, message, compiledScript,
                     instanceActor, sharedScriptLibrary, options, completionTarget, logger,
                     healthCollector, serviceProvider, runId, parentExecutionId, timeout,
-                    executionScheduler, cts),
+                    executionScheduler, cts, siteCommunicationActor),
                 CancellationToken.None, TaskCreationOptions.DenyChildAttach, executionScheduler)
                 .Unwrap();
         }
@@ -659,7 +661,8 @@ internal static class ScriptRunLauncher
         Guid? parentExecutionId,
         TimeSpan timeout,
         ScriptExecutionScheduler executionScheduler,
-        CancellationTokenSource cts)
+        CancellationTokenSource cts,
+        ICanTell? siteCommunicationActor)
     {
         // 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
@@ -696,6 +699,9 @@ internal static class ScriptRunLauncher
             // run it late — the alarm it would react to is already stale.
             cts.Token.ThrowIfCancellationRequested();
 
+            using var serviceScope = serviceProvider?.CreateScope();
+            var services = serviceScope?.ServiceProvider;
+
             // An alarm on-trigger run can call Instance.CallScript()
             // via the ScriptRuntimeContext injected into globals
             var context = new ScriptRuntimeContext(
@@ -711,6 +717,15 @@ internal static class ScriptRunLauncher
                 // it every Log.* line from an alarm script was attributed to the INSTANCE, not the
                 // alarm. `scriptName` above is already the "AlarmActor:{alarmName}" source key, so
                 // it is passed as-is (re-prefixing it would leave "AlarmActor:" in the target).
+                externalSystemClient: services?.GetService<IExternalSystemClient>(),
+                databaseGateway: services?.GetService<IDatabaseGateway>(),
+                storeAndForward: services?.GetService<StoreAndForwardService>(),
+                siteCommunicationActor: siteCommunicationActor,
+                siteId: services?.GetService<ISiteIdentityProvider>()?.SiteId ?? string.Empty,
+                auditWriter: services?.GetService<IAuditWriter>(),
+                operationTrackingStore: services?.GetService<IOperationTrackingStore>(),
+                cachedForwarder: services?.GetService<ICachedCallTelemetryForwarder>(),
+                sourceNode: services?.GetService<INodeIdentityProvider>()?.NodeName,
                 sourceScript: scriptName,
                 // ParentExecutionId tag-cascade: the
                 // alarm on-trigger run mints its own fresh ExecutionId (the