A run is one execution of a saved workflow version. Each step inside it records its own status and outcome, which is what makes a failure something you can read rather than reconstruct.
Choose your next task
- Watch runs in Monitor — the run history, its filters, live tail, and what a run's detail view shows.
- Recover a failed run — read the failure, retry from the failing step, and resolve calls whose outcome is unknown.
- Activation errors — for a workflow that refuses to go live, which is a different problem from a run that failed.
Where to start when something is wrong
- Did it run at all? If no run appears, the problem is upstream of execution — the workflow may be inactive, or the trigger never fired. See Save, version, and activate.
- Did the run fail? Open it and find the first step with status failed. The error is on that row.
- Was the failure transient? Retry from the failure. Every step that finished is kept.
- Was it a configuration problem? Fix the workflow, save, and activate the new version — then retry, or wait for the next trigger.
Saving builder changes and activating are separate actions, so before testing a fix, confirm which saved version is actually active.
Related
- Build your first workflow if you have not produced a run yet.
- Retries to stop transient failures becoming failed runs in the first place.