See llms.txt for all machine-readable content.
@jgivati
I work on n8n reliability, specifically the failures that report success. A workflow that runs clean and produces nothing throws no error, so the error workflow never fires and nobody finds out until someone downstream asks where the data went. My templates come out of that work. They check an instance from outside, through the n8n API, so they keep reporting even when the instance itself is what broke. ranfine.io