Jobs and Schedules

Jobs are durable server-only work. Schedules create ordinary Privileged Jobs at declared times; they do not introduce a second execution system.

Declare a named job() when work must outlive the request that created it. Enqueue as the captured current user when later ACL checks should retain that identity; use the Privileged server role only for system-owned work. Handlers must tolerate duplicate execution because delivery is at least once.

Declare schedule() only for recurrence. Pin an IANA timezone when wall-clock meaning must remain consistent across Dev, Container, and Hosted environments. Schedules enqueue Jobs; retry and execution semantics remain Job Queue semantics.

Read the Job Queue and Schedule walkthrough, then use the actor-selection guidance to choose a captured current user or the Privileged server role.

Use sporades jobs and sporades schedules for bounded JSON inspection of an active Dev session; equivalent deploy and host commands inspect other environments.