I’ve been building Quorum, an open-source tool for people running important n8n workflows.
You tell Quorum what “working” means for a workflow. For example, it should run every 15 minutes, or a successful run should return at least one useful item. Quorum then checks whether that actually happened from outside n8n.
The main reason I built it is that some failures never create an execution at all. A workflow gets deactivated, a schedule stops firing, or the instance is down at the wrong time. There is nothing in the execution history because the run never existed.
When Quorum detects that, it opens one incident, keeps updating it while the problem continues, and resolves it when the workflow starts behaving normally again.
It’s self-hosted, open source, and still early. I’m looking for people running real n8n workflows who are willing to connect one and tell me where the setup is confusing or where the idea falls apart.
Demo: https://quorumwatch.com
GitHub: https://github.com/werniq/quorum_core
How do you currently notice when a workflow simply never starts?