Error handling

Continue on error (Power Mode)

Any action step can be set to continue on error. If it fails, the run doesn't stop — the failed step shows an amber warning badge with a "Last Run Error" card (message, timestamp, and a suggested fix where Cheqq can identify one, like a Reconnect button on an auth failure), and downstream steps still run.

Atomic mode

A workflow can opt into atomic execution. When it's on, if any step fails partway through a run, Cheqq automatically undoes the collection record changes made by earlier steps in that same run — in reverse order (an insert is undone with a delete, an update is undone by restoring the prior value, a delete is undone by restoring the record).

Test Step, Dry run, and Run now

• Test Step, from a step's detail panel, runs just that one step in isolation — useful while wiring up a new action without running the whole workflow.
• Dry run executes the whole workflow, but substitutes safe simulated output for any step that would otherwise have a real side effect — API call, integration action, AI call, CLI, notification, send email, and RSS read are all stubbed. Control-flow steps (Branch, loops, Parallel) and Tool/Code steps still execute for real, so you can confirm your branching logic actually takes the path you expect, without anything external actually happening.
• Run now performs a full, real run.

While a run is in progress the canvas locks — nodes, edges, and step configuration can't be edited — and the toolbar shows Stop instead. Stopping is cooperative: the current step finishes, and no further steps start.

Pausing a live run

Separately from dry-run, an in-progress run can be paused — it checkpoints between steps (never mid-step) and can be resumed later from where it left off.

Continue with Step types or Triggers, or see Troubleshooting for failed runs.