Automation

Schedules

A Schedule is a recurring automation. Like cron jobs in a Unix system, Schedules run in the background on a cadence without you having to think about them. "Every Monday at 9am, do this." "On the first of every month, generate that."

In River, a Schedule is a first-class entity — you can see it, edit it, pause it, and delete it just like any other thing in your workspace. Each Schedule fires a Job on its cadence.

What Schedules can fire#

Schedules don't care what kind of Job they fire. They can run:

  • An Agent (the most common case)
  • A Workflow
  • A Tool call directly (e.g., a daily database query)
  • A Skill (a single LLM call)

Pick the cadence; pick what runs. That's it.

Common cadences#

You can use plain English or cron syntax:

Plain Cron
Every weekday at 7am 0 7 * * 1-5
Every Monday at 9am 0 9 * * 1
Every hour 0 * * * *
The first of each month 0 9 1 * *
Every 15 minutes */15 * * * *

The Schedule editor accepts both and shows the next 5 fire times so you can sanity-check.

Where Schedules live#

Open AutomationsApp → Schedules to see all Schedules in your Dataspace, their next run time, and the last result. From there you can:

  • Pause or resume a Schedule
  • Edit the cadence
  • Run it now (independent of its schedule)
  • Look at the Job history (every past run)

Common patterns#

  • Morning briefing at 7am weekdays — runs your Morning Briefer Agent.
  • Weekly recap Friday at 4pm — runs a "Last Week in Review" Agent.
  • Monthly close the first of the month — runs a Workflow that pulls Stripe data and generates a financial summary.
  • Hourly inbox sweep during business hours — runs an Inbox Triage Agent on emails received in the last hour.

Quiet hours and weekends#

You can scope a Schedule to:

  • Specific days (weekdays only, weekends only, custom)
  • Specific hours (business hours only)
  • Specific time zones (the user's local time, not UTC)

A Schedule paused for the weekend will skip Saturday and Sunday automatically.

Tips#

  • Always check the "next run" preview before saving — it catches obvious mistakes (wrong time zone, wrong day).
  • Pause Schedules instead of deleting them while iterating. Less likely to lose history.
  • Schedules in a team Dataspace are shared — anyone can see what's running and when.