v0.52.7
July 19, 2026
An engine that holds together even under failure: if external automation is unreachable, tasks fail cleanly — with alerts and fallback paths — instead of staying "running" forever; lists and dashboards are lighter; and every operation can be traced in the logs from click to webhook.
-
Improvement
Failures handled, not suffered
If n8n or an external service is down, webhook tasks no longer stay stuck "Running" indefinitely: once retries are exhausted, the task is marked Failed with details in the audit trail, your configured alerts fire and the flow's fallback rules kick in. And while the service is unreachable, new deliveries wait a minute instead of hammering a dead endpoint. Alerts are fairer too: no backlog can hide newly waiting items, and a superseded alert no longer silences the item-level ones.
-
Improvement
Lighter lists and dashboards
The task list is now server-side paginated with a total and an indicator when there's more than shown; automation KPIs and version comparison no longer run one query per flow; Process Map insights no longer saturate the database. Same numbers, faster responses, less load.
-
Improvement
Operations traceable end to end
Every job started by a request (triggers, events, transitions) carries the request identifier and worker logs report it: from the user's click to the delivered webhook, a single thread to follow. And service health is honest: the worker truly reports when the queue engine has a problem.