Make webhook logs kept 3 days: recover a Sume result later
Make keeps webhook logs for 3 days (30 on Enterprise). A Sume result does not depend on them: store the job id, read job events, and redeliver the callback.

Once Make's webhook log is gone, the Sume job is not. Make's help page says it stores webhook logs for 3 days, or 30 days on Enterprise. A Sume job keeps its own record, so the way back is the stored job_id: read the job and its events, and if the callback never landed, call redeliver.
Make's retention is from its Webhooks help page; the Sume side is from Jobs and results and Webhooks, read 2026-10-01.
What should I store when the scenario submits?
Sume's docs say to store the job id from submit responses so your integration can recover work after process restarts. In Make, write it to a data store or a sheet row next to the request, not only to the scenario's execution history.
Where do I see what happened to the callback?
Job events are a public timeline for debugging and recovery, and they include webhook.delivery. The delivery status vocabulary is pending, delivering, delivered, retrying, failed, exhausted; the attempt count is visible on the job object and in job events when available.
| Question | Where to look |
|---|---|
| Did the job finish? | Job status via status_url |
| Was a callback attempted? | webhook.delivery job events |
| Is the result still readable? | result_url once result_ready is true |
| Need the callback again? | POST /v1/jobs/{job_id}/webhook/redeliver |
How do I get the callback again?
Redeliver re-POSTs the job's real terminal event (job.completed, job.failed or job.canceled) with a fresh timestamp and signature, and the docs say this still works after automatic attempts are exhausted. It needs the jobs:write scope and does not change the destination URL.
Do not use Send test for this: it posts a dummy webhook.test payload and never replays a real job.
curl -X POST https://api.sume.com/v1/jobs/job_123/webhook/redeliver \
-H "Authorization: Bearer $SUME_API_KEY"Will redelivery run my scenario twice?
It can, if the first delivery did arrive. Treat job_id as the idempotency key on the Make side. For more on failed deliveries see debug failed webhook deliveries.
Sources
Related posts
- Can an AI agent debug a failed webhook on Sume over MCP?
- Make webhook "Queue is full" 400: what Sume does and how to recover
- Make.com AI video scenario with Sume: HTTP and a webhook
- AI music API webhook: get a callback when the track is ready
- Video generation API timeouts: Sume's wait caps, SDK defaults, expiry
More in Integrations
- Make webhook queue MCP tools: stuck items and Sume redelivery
When a webhook queue item looks stuck, check Make's queue and Sume's delivery status separately. Sume can re-POST a job's terminal event with Redeliver.
- Mastra idleTimeoutMs observer detach: the Sume job keeps running
Mastra idleTimeoutMs detaches the observer and keeps the run. A Sume job it started keeps running too: re-wait on its id, do not resubmit.
- Mastra respondToToolApproval needs toolCallId: Sume paid calls
Mastra now rejects approval responses without toolCallId. When you approve a Sume paid tool call, key it to that id and keep its idempotency_key stable.
- MCP Enterprise-Managed Authorization is stable: Sume uses OAuth or key
The MCP roadmap calls Enterprise-Managed Authorization stable. Sume's hosted MCP documents two sign-in modes: OAuth with scopes, or an API key.
Written by Sume