Good point, that’s a better approach. I guess for a multi-repo situation at work, you would need to create a base project like “postgres-lock-ids” so you can synchronize the lock across everything.