feat: release 2.2.0 with PC-only tests and execution reliability fixes

Unify task scheduling and queue state, fix Android input and observation recovery, preserve screenshot paths, and match final review criteria accurately. Add collapsible copyable results and publish the complete 2.1.0 upgrade notes. Validation: 3189 backend tests, 326 frontend tests, 24 client tests, production build and quality gates.

apps/admin_console/services/task_queue_service.py

@@ -346,7 +346,7 @@ class TaskQueueService:
 
     @classmethod
     def _concurrency_limit(cls) -> int:
-        """Return 0 for per-device concurrency, or the global task limit."""
+        """Return the shared global single-task limit, including computer targets."""
         return DeviceExecutionLock.resolve_env_concurrency()
 
     @classmethod
@@ -354,10 +354,8 @@ class TaskQueueService:
         """Persistent dispatcher scheduling pending tasks onto devices.
 
         Scans the queue and launches each eligible task as an independent
-        coroutine. Admission is governed by _concurrency_limit(): 0 admits one
-        task per device so distinct devices execute concurrently, N>=1 admits
-        at most N tasks across all devices. Per-device FIFO ordering and
-        cross-process mutual exclusion remain enforced by DeviceExecutionLock
+        coroutine. Admission allows one executing task across all targets.
+        Global FIFO ordering and cross-process mutual exclusion are enforced by DeviceExecutionLock
         inside each worker process.
         """
         print(
@@ -436,6 +434,8 @@ class TaskQueueService:
             target = cls._task_target(item)
             resources = set(item.get("plugin_resources", []))
             if resources & busy_resources:
+                if limit == 1:
+                    return
                 continue
             if limit == 0:
                 # A task without a resolved device may bind to any serial, so it
@@ -451,6 +451,8 @@ class TaskQueueService:
                     continue
             elif limit >= 1 and device is not None and target.lock_key in busy_devices:
                 # A second worker for this device would wait on its lock.
+                if limit == 1:
+                    return
                 continue
 
             item["status"] = "running"