Check on automations
With automations:read:
| Call | Returns |
|---|---|
GET /v1/automations?project_id=PROJECT_ID | Automations you can see, each with its trigger and a one-sentence summary |
GET /v1/automations/AUTOMATION_ID | One automation |
GET /v1/automations/AUTOMATION_ID/runs | Its recent runs, newest first |
GET /v1/automation-runs/RUN_ID | One run, with every step it took and where it stopped or failed |
A failed run has an error, and its steps show which one failed. That makes a simple monitor easy: list recent runs every few minutes and alert on new failures.
Start one from your code
Automations whose trigger is Started by hand have startable: true. These can be started with automations:write, for example from a form submission or another system's webhook.
- 1
Start it
curl -X POST https://developers.playbookops.ai/v1/automations/AUTOMATION_ID/runs \ -H "Authorization: Bearer $PLAYBOOKOPS_API_KEY" -H "Content-Type: application/json" \ -H "Idempotency-Key: $(uuidgen)" \ -d '{"item_id": "ITEM_ID"}'If the automation has
needs_item: true, pass theitem_idto run it on. An app's automation takes arecord_idinstead. - 2
Follow the run
The answer is the new run. Check on it with
GET /v1/automation-runs/RUN_IDuntil it has afinished_at, then look at itsstatus, itserrorif it failed, and its steps.
The automation must be published and switched on. Automations with any other trigger start on their own, and starting them returns 400.
Reference
| Action | Endpoint | MCP tool |
|---|---|---|
| List automations | GET /v1/automations | list_automations |
| One automation | GET /v1/automations/{id} | get_automation |
| Its runs | GET /v1/automations/{id}/runs | list_automation_runs |
| One run | GET /v1/automation-runs/{id} | get_automation_run |
| Start one | POST /v1/automations/{id}/runs | start_automation |
Every parameter is in the API reference.