Schedule node and scheduled jobs

Use a normal schedule when timing and inputs are known in advance. Use a Schedule node when another workflow must determine those inputs at runtime.

Choose the right approach#

A scheduled worker does not need a Schedule node inside itself. The schedule starts that worker.

Configure a Schedule node#

Select the target action, schedule timing, state, and body. The body can use values from preceding nodes. The target receives that stored body as ${body} when it executes.

{
  "orderId": "${CreatedOrderId}"
}

The node creates or updates a schedule using the target action, normalized body, and environment. It does not execute the target inline. A one-time active schedule created by the node is queued after the initiating transaction commits.

Review recurring timing#

Choose an interval supported by the schedule editor. Daily, weekly, and monthly settings use a UTC start time; confirm the displayed timezone before translating a business deadline. Review the weekday or day-of-month selection and the schedule's active state.

Make jobs safe to repeat#

Jobs can be retried or manually replayed. Use a stable business identifier and make repeated runs detect already-completed work. Put required data in the schedule body or a database, not only in the original run's session keys.

Inspect schedule state and resulting action logs when a job does not run. Confirm its environment, target action, release availability, and inputs. See the scheduled worker guide for a complete setup.