fix: bound task failures and release stalled workers

apps/admin_console/services/automation_service.py

@@ -10,6 +10,8 @@ from uuid import UUID
 from apps.admin_console.core.state import state
 from apps.admin_console.services.automation_store import store, new_id
 from apps.admin_console.services.task_import import case_goal, case_title
+from artemis.config.settings import settings
+from artemis.runtime.task_limits import task_timeout
 
 logger = logging.getLogger(__name__)
 TERMINAL = {"completed", "failed", "cancelled", "skipped", "interrupted"}
@@ -135,10 +137,11 @@ async def tick_batches():
                 # not consume this case's execution budget.
                 if job["status"] != "running":
                     continue
-                limit = int(
-                    (job.get("case", {}).get("task_config") or {}).get("timeout_seconds")
-                    or batch.get("timeout_seconds", 1800)
-                )
+                from artemis.runtime.task_result import read_worker_outcome
+
+                if read_worker_outcome(str(job["session_id"])):
+                    continue  # Execution has ended; the worker supervisor bounds cleanup.
+                limit = task_timeout({}, job, batch)
                 task_state = getattr(state, "active_runs", {}).get(str(job["session_id"])) or {}
                 excluded = float(task_state.get("non_execution_seconds") or 0)
                 if task_state.get("non_execution_started_at"):
@@ -151,7 +154,17 @@ async def tick_batches():
                         error=f"任务执行超时（{limit} 秒，不含暂停/排队时间）", timed_out=True
                     )
                     store.save_job(job)
-                    task_queue_service.stop_tasks(session_id=job["session_id"])
+                    from artemis.runtime.task_result import save_failure_result
+
+                    try:
+                        await asyncio.to_thread(
+                            save_failure_result,
+                            str(job["session_id"]),
+                            job["error"],
+                            code="task_timeout",
+                        )
+                    finally:
+                        task_queue_service.stop_tasks(session_id=job["session_id"])
                     # on_terminal records the failed case. The next pending
                     # case then runs normally unless stop_on_failure is set.
             continue
@@ -255,8 +268,7 @@ def create_batch(payload: dict) -> dict:
         "formats": payload.get("formats") or store.preferences()["formats"],
         "profile": payload.get("profile", "flash"),
         "device_serial": payload.get("device_serial"),
-        "timeout_seconds": payload.get("timeout_seconds")
-        or int(os.getenv("ARTEMIS_BATCH_TASK_TIMEOUT", "1800")),
+        "timeout_seconds": payload.get("timeout_seconds") or settings.ARTEMIS_TASK_TIMEOUT_SECONDS,
         "stop_on_failure": payload.get("stop_on_failure", False),
         "request_id": payload.get("request_id"),
     }