fix(engine): dispatch run continuation after releasing the lock

Prevents a second worker from consuming the follow-up job and bailing on the
still-held run lock, which would stall the run until the next scheduler tick.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

app/Provisioning/RunRunner.php

@@ -18,8 +18,12 @@ class RunRunner
 {
     public function __construct(private PipelineRegistry $registry) {}
 
+    private ?string $pendingContinuation = null;
+
     public function advance(ProvisioningRun $run): void
     {
+        $this->pendingContinuation = null;
+
         // Lock must outlive the longest step so a duplicate job (dispatched by
         // the minutely tick while this one still runs) can't execute concurrently.
         $lock = Cache::lock('run:'.$run->uuid, 2100);
@@ -33,6 +37,13 @@ class RunRunner
         } finally {
             $lock->release();
         }
+
+        // Dispatch the continuation only AFTER releasing the lock, so a second
+        // worker can't consume it and then bail on the still-held lock.
+        if ($this->pendingContinuation !== null) {
+            AdvanceRunJob::dispatch($this->pendingContinuation);
+            $this->pendingContinuation = null;
+        }
     }
 
     private function runLocked(ProvisioningRun $run): void
@@ -116,7 +127,7 @@ class RunRunner
         $run->save();
         $this->record($run, $step->key(), 'advanced', null);
 
-        AdvanceRunJob::dispatch($run->uuid);
+        $this->pendingContinuation = $run->uuid; // dispatched after the lock releases
     }
 
     private function onRetry(ProvisioningRun $run, ProvisioningStep $step, StepResult $result): void