Scheduling

Modes

• cron: 5-field cron (minute hour day month weekday) in a configured timezone.
• interval: fixed interval in seconds (interval_seconds).
• one_time: executes once at one_time_at.
• manual: no automatic execution; trigger using POST /schedules/{schedule_id}/trigger.

Shared attributes

• timezone: IANA timezone (default America/New_York for implicit schedules created via deploy).
• enabled: include/exclude from automated scheduling.
• start_at / end_at: optional scheduling boundaries.
• catchup_policy:
  • none: skip backlog and schedule from current time.
  • all: keep advancing from previously due time until backlog is drained.

Implicit schedules via deploy

When deploying a flow with a schedule parameter (via flow.deploy() or dagy deploy --schedule), an implicit cron schedule is automatically created. You can specify the timezone for this schedule:

Python SDK:

flow.deploy(
    name="my-pipeline",
    schedule="0 * * * *",
    timezone="America/Los_Angeles",
)

CLI:

dagy deploy artifact.zip --deployment my-pipeline --schedule "0 * * * *" --timezone "America/Los_Angeles"

If no timezone is provided, it defaults to America/New_York.

Runtime fields

• next_run_epoch / next_run_at: next due execution.
• last_triggered_at: most recent scheduler/manual trigger timestamp.
• last_run_id / last_run_slug: most recent run linkage.
• last_error: latest scheduler-trigger error.