Skip to content
API Reference

Inspect Workflow Runs and Troubleshoot Failures

Version: unreleased · Docs: 3c6ab9f2Download Markdown

Inspect the run rather than relying only on the workflow’s saved configuration. Successful earlier steps can coexist with a failed final action.

Check a run at each stage: A later failure can follow a successful write; inspect the destination before retrying.

A workflow may finish several steps before one fails. Start at the first failed or waiting step, inspect the value it received, and compare it with that step’s required fields. Later steps can have no output simply because the earlier dependency never completed.

For an external write, check whether the target was changed before retrying. A file may have been created even if a later notification failed. Rerunning the entire flow without checking can create another file or send another message. Use run history to distinguish separate attempts.

  1. Open the workflow’s run view and relevant run timeline.
  2. Confirm the supplied inputs and the trigger that started it.
  3. Expand step activity to inspect outputs and errors.
  4. Resolve any pending question or approval.
  5. Confirm the final output and any destination-app changes.

For a checklist workflow, compare owners and deadlines with the input. For an app action, follow the returned file or message reference and inspect the result in that app.

Situation Next check
Missing required input Complete the run form or upstream binding
Waiting for approval Review the exact proposed action
Permission or not-found error Check account, resource ID, and access
Worker or sandbox interruption Check runtime availability with an administrator
Unknown write outcome Inspect the destination before retrying

Correct the cause before repeating a run. Retrying a workflow can repeat earlier writes even if a later step was the only failure. Use read-only checks or a small input while diagnosing an external action.

The workflow actions menu offers Analytics for the workflow’s available run analysis. Check automatic triggers if failures recur without a manual start. Save corrected configuration before another attempt.

Related: triggers and app actions.