Skip to content

Run and monitor a workflow

  • Debug run, from the editor, executes immediately and shows each task’s input, output and error as it happens. Use it while building.
  • Run, from the dashboard, starts a normal run.
  • Triggers start runs without you: a schedule, or an incoming webhook.

The Runs tab lists every run with its status.

Status Meaning
Pending Created, not started yet.
Queued Waiting for a free slot to start.
Running Executing.
Suspended Waiting for something outside the run — an approval, an input, or a retry.
Succeeded Finished; the result is recorded.
Failed Stopped on an error; the error is recorded.
Retry pending A failed step is waiting to be tried again.
Dead letter Retries are exhausted; the run will not continue on its own.

Open a run to see every task, what it received and what it produced. The same view works while the run is still going.

A debug session that sits idle pauses itself, so leaving the editor open does not keep spending credits. A run that suspends — waiting for an approval or an input — resumes when the answer arrives and keeps the progress it had made.

A run is charged for the time it runs: one credit per second, at least one credit, however it was started — a schedule, a webhook, the Run button or a debug session. The charge does not depend on what the workflow does, and a run that fails or is stopped is charged for the time it ran.

Two runs you start yourself can be in flight at once; a third is refused while the other two are running. Scheduled and webhook runs do not take one of those slots. A single attempt is stopped after 15 minutes and marked failed, and a resumed run starts a new attempt with a fresh 15 minutes. The plans page lists what each plan includes.