Module · Executions & debugging✅Knowledge checkCheck: finding out what really happened4 questions · Pass at 60%1. You fix a node's config and press Run From Here. What is it reusing, and why is that both the point and the trap?The earlier nodes' existing results — cheap, and possibly staleThat node's previous output, so you can compare before and afterNothing at all; it re-runs the whole workflow from the trigger downThe last successful execution's inputs, which are always current2. A colleague asks you to send a link to a specific execution. What can you do?Copy the URL, since selecting a run updates the page's addressShare the workflow link with the execution id as a query parameterNothing — the address does not change, so send an identifierExport the run from the panel, which produces a shareable link3. You are asking the AI debugger why one node produced an empty field. What gets you a better answer?Handing over the whole run, so it has all the context it might needHanding over the workflow definition as well as the executionAsking generally, so that it is not led toward a wrong conclusionScoping it to that node — widening makes answers drift4. What does “testing a workflow” mean, after twenty lessons of evidence?Confirming that the run completed with no failed nodes in itReading what the nodes actually produced, on realistic dataRunning each node on its own until every one of them passesChecking that the validator reports the workflow as runnableCheck answers← Back to Webhooks, Debugging & Full BuildsNext topic →
