Automation Did Not Start or Is Still Waiting

Diagnose why an automation failed to trigger or paused on a waiting Step. Start by deciding which of those two states you have.

At a glanceVerified in app · Oct 30, 2026

Use it to: Diagnose why a trigger did not start a run or why an active run is paused on a Wait or Human Review Step
Find it: Canvas → Live toggle, Trigger App drawer, and Execution Log
Limits: 10 MiB default incoming webhook payload · Up to 13 days Wait timer · Up to 30 days Human Review expiry
Instead: Debug a Run (when a run started and completed or failed with unexpected results)

Automation did not start

First confirm whether the Execution Log contains a run for the expected time or event.

No run exists: check that the automation is Live. Draft allows manual runs but disables automatic triggers. For a Scheduler, also check the timezone and the next run times shown in its panel. A scheduled occurrence that did not start is not replayed later.
A run exists: check its origin and input to confirm it belongs to the expected event, rather than a manual test or another trigger. Then find the first failed, skipped, or unexpected Step.

An app trigger needs attention

Open the trigger’s App drawer and read its status. Reconnecting is not the remedy for every paused trigger:

Status	What to do
Needs reconnecting	Ask the connected account’s owner to reconnect it. Glow lacks a working sign-in for this trigger.
Your admin needs to finish app setup	Ask a workspace admin to configure how the app sends events to Glow. There is nothing to change in the trigger itself.
Bot event setup required	Ask a workspace admin to open the bot connection’s details and finish event delivery setup.

After the change, check the trigger’s status again and confirm that a controlled event starts a run.

A webhook delivery did not start a run

Check the sender’s delivery log, then compare it with the Webhook trigger’s current state:

Confirm the automation is Live. A webhook URL can remain available in Draft or while paused without starting runs.
Copy the current Webhook endpoint from the trigger. The default Webhook response is 200 {"success": true}, but the response is configurable. A 202 {"received": true} can also be returned when no event is accepted for execution. Neither response proves a run started or completed.
If signature verification is enabled, confirm the sender uses the configured secret and header. A live trigger rejects an unsigned or wrongly signed request before a run exists.
Check the configured incoming payload limit. The default is 10 MiB, not the separate 1 MiB response-body cap. Confirm deployment-specific limits before sending large requests; see Payloads and storage.
Send one controlled request and look for it in the trigger’s Executions tab. Review downstream actions first: the request can start real writes and consume usage.

Use the Execution Log and connected services to confirm the outcome before resending. A successful HTTP acknowledgement is not a completion receipt or duplicate protection.

POST to your webhook URL

{
  "customer_name": "Acme Corp",
  "amount": 4500
}

Typeform, Stripe, GitHub or your own script sends JSON to the URL.

The run is waiting

Open the waiting Step’s Executions tab and inspect its continuation details. Downstream Steps do not run until the wait resumes or expires:

Waiting step or mode	How it resumes	Draft vs Live and expiry behavior
Wait (timer)	Resumes automatically when the configured duration elapses (up to 13 days).	Recorded before the run pauses.
Wait (On webhook call / On form submitted)	Caller invokes the single-use `resumeUrl` or `formUrl` generated for that specific run.	`resumeUrl` requires the automation to remain Live; a hosted `formUrl` can also resume an active manual run in Draft.
Human Review	Reviewer submits a decision on the review request (expiry up to 30 days; `0` or blank waits without expiry).	Routes to the Expired branch if connected; otherwise follows the Step’s error path.

Verify it works

Run: Trigger one controlled event or resume the waiting Step, then open the Execution Log.
You should see: A new run starts from the trigger, or the paused run advances past the waiting Step.

What’s Next?

Trace an existing run with Debug a Run.
Choose the correct trigger in Triggers and Actions.
Inspect completed and active runs in the Execution Log.