An automation that runs on every incoming request needs to be right before the first real request arrives. That is what Test run is for.
What a test run does
Test run sends the sample JSON in the Test input box through the workflow exactly as a webhook request would. Every step runs, conditions pick a branch, and actions are attempted. The result is saved in Run history with the label "test run" so you can tell it apart from live traffic.
Reading the run history
Each run lists every step in order with a status: succeeded, skipped or failed, plus a short message. A skipped Slack step usually means Slack is not connected yet; a failed webhook action shows the status code the other system returned. The AI step's message shows the summary, priority and category it produced, which is the quickest way to see why a request took the branch it did.
A checklist before you share the URL
- Run one test with an urgent sample and one with an ordinary one, and check each took the branch you expected.
- Check every action shows succeeded, not skipped. Connect anything that is missing on the Integrations page.
- Read the Slack or email message the run produced. Placeholders like
{{summary}}should be filled in, not shown literally. - Send one real request from the tool that will call the workflow, and confirm it appears in Run history as a webhook run.
After that, the run history becomes your log: when someone asks why a request was or was not flagged, the answer is one click away.