Skip to main content
Every time a workflow runs — from a manual click, a schedule, a webhook, or any other trigger — an execution is recorded. The Executions tab of the workflow editor shows this history, so you always know what ran, when, and how it went.

The Executions tab

Open a workflow and switch from Editor to Executions. Each run appears with its status, duration, and timing. If the workflow has never run, the tab shows an empty state: “No executions yet — Run this workflow to see executions here.”

Inspecting a run

Select a run to open it. You can walk through the workflow node by node and see the output each node produced — the data it received, what it did with it, and what it passed along. On failed runs, the node that errored shows the error itself, so you can see exactly where and why the run stopped. This makes executions the main debugging surface for workflows: instead of guessing, you look at real data from a real run.

Debug a failing workflow

1

Run the workflow

Select Execute Workflow in the editor to produce a fresh run.
2

Open the failed run

Switch to the Executions tab and select the run marked Error.
3

Inspect node outputs top-down

Follow the run from the trigger downward. Check each node’s output until you reach the node that failed, and read its error. Often the real problem is one node earlier — a previous node handing over data in a shape the failing node didn’t expect.
4

Fix the node configuration

Go back to the Editor tab and adjust the configuration of the node at fault.
5

Save

Select Save. This step matters — see the tip below.
6

Run again

Select Execute Workflow once more and confirm the new run succeeds.
Runs always use the last saved version of the workflow. If you fix a node but forget to Save, the next run repeats the old failure — and it’s easy to mistake that for the fix not working. When a run doesn’t reflect your change, check the last-saved time in the header first.

Workflows

The editor, the workflow list, and building your first automation.

Node reference

What every execution node does — useful when reading a run’s outputs.