Auditing an n8n workflow you didn't build
You inherited a flow from a client or a former teammate. It runs, but you have no idea what it drops when something goes wrong.
Why it happens
Most hand-built automations skip reliability basics: webhook ordering, error branches, retries and dedupe. Each gap is an invisible leak.
How to fix it
Check four things: does the webhook respond after the save; does every failable node have error handling; do HTTP calls retry; and is there dedupe before writes.
How to verify
Run a test record end-to-end and confirm it survives a forced failure at each step.
Not sure if this is the only leak?
Paste the export and get the full leak map in 30 seconds — no account, nothing stored.
Scan my workflow freeRelated leaks