Skip to main content
Schedules is a legacy feature — new scheduled work is created as an automation. Automations support schedule triggers along with event-driven triggers (Slack, GitHub, Linear, webhooks), conditions, invocation limits, and more. To run Devin on a schedule, open Automations in the sidebar, click Create automation, and add a Schedule trigger. Existing scheduled sessions will continue to work, and Devin can help migrate them to automations — just ask.
Scheduled Sessions let you create Devin sessions that run automatically — either on a recurring schedule or as a one-time run at a specific date and time. Use them to automate repetitive tasks like daily reports, periodic code maintenance, routine data analysis, and more.

​
Creating a Scheduled Session

New scheduled sessions are created as automations with a Schedule trigger. There are two ways to start:

​
From the automations page

  1. Navigate to Automations in the sidebar
  2. Click Create automation
  3. In the trigger dropdown, expand Schedule and choose Every hour, Every day, Every week, Run once, or Custom schedule
  4. Write the prompt Devin should follow on each run, then click Save

​
From an existing session

  1. Open the menu on a session in the sidebar
  2. Select Schedule Devin
  3. You’ll be taken to the automation editor with that session’s prompt pre-filled
See Schedule triggers for the full set of scheduling options. The rest of this page describes the options available on existing legacy schedules.

​
Configuring a Schedule

When editing an existing schedule, you can configure the following options:

​
Name

Give your schedule a descriptive name so you can easily identify it in the list (e.g., “Daily CI Report” or “Weekly Dependency Updates”).

​
Schedule type

Choose between two schedule types:
  • Recurring — Runs repeatedly on a cron-based frequency (default)
  • One-time — Runs once at a specific date and time, then automatically disables itself

​
Agent

Choose which agent type should run the scheduled session:
  • Devin — Standard AI software engineer (default)
  • Data Analyst — Optimized for data analysis and queries
  • Advanced — For playbooks and session analysis

​
Playbook (optional)

Attach a playbook to the scheduled session. The playbook will be applied every time the schedule runs, ensuring consistent behavior across executions.

​
Repositories (optional)

Select one or more repositories for the scheduled session to work in. When repositories are selected, they are included as a hint in the session prompt so Devin knows which repos to focus on. Leave empty to let Devin determine the relevant repositories from the prompt.

​
Frequency (recurring schedules)

For recurring schedules, set how often the schedule should run. The frequency editor supports two modes: Visual mode provides preset options:
  • Hourly — Run every N hours
  • Daily — Run at a specific time every day
  • Weekly — Run at a specific time on selected days of the week
Times are displayed in your local timezone but stored as UTC internally. The editor handles the conversion automatically.
Custom mode lets you enter a standard cron expression directly (e.g., 0 9 * * 1-5 for weekdays at 9 AM UTC). This gives you full flexibility for complex schedules.

​
Run at (one-time schedules)

For one-time schedules, pick the date and time when the session should run. The time is entered in your local timezone and converted to UTC automatically. One-time schedules must be set to a time in the future. After a one-time schedule executes, it is automatically disabled. The schedule and its past sessions are preserved for audit purposes.

​
Email notifications

Control when you receive email notifications about scheduled session runs:
  • Always — Get notified after every run
  • On failure only — Only get notified when a scheduled session fails (default)
  • Never — No notifications

​
Slack notifications

You can optionally send notifications about scheduled session runs to a Slack channel. When configured, Devin posts updates to the selected channel when the schedule runs. This requires a connected Slack integration in your organization settings.

​
Run as

By default, sessions created by a schedule are attributed to the user who originally created the schedule. When editing a schedule, you can update this so that future runs are attributed to you instead. This is useful when schedule ownership changes hands — the new owner receives notifications and the sessions appear under their account.

​
Prompt

Write the instructions that Devin will follow each time the schedule runs. This is the same as the prompt you would type when starting a regular Devin session.

​
Managing Schedules

Existing schedules don’t have a sidebar entry. Open app.devin.ai/settings/schedules to see all your legacy scheduled sessions. The list shows each schedule’s name, frequency, last run time, and status.

​
Status

Each schedule has one of three statuses:
  • Active — The schedule is enabled and will run at its next scheduled time
  • Paused — The schedule is disabled and will not run until re-enabled. One-time schedules are automatically paused after execution.
  • Error — The schedule encountered consecutive failures

​
Editing a schedule

Click on any schedule in the list to view its details. Click Edit to modify its configuration, including the name, prompt, agent, playbook, repositories, frequency, notification settings, run-as user, and whether it is enabled or paused.

​
Pausing and resuming

You can pause a schedule by editing it and toggling the Status switch to Paused. Paused schedules will not create new sessions until re-enabled. Toggle it back to Active to resume.

​
Deleting a schedule

Click the three-dot menu on a schedule’s detail page and select Delete. This permanently removes the schedule. Past sessions created by the schedule are not affected.

​
Viewing Past Sessions

Each schedule detail page has a Past Sessions tab that lists all Devin sessions created by that schedule. Click on any session to navigate to its full session view. This is useful for reviewing outcomes, debugging failures, or auditing what the schedule has been doing over time.

​
Use Cases

Here are some common ways to use Scheduled Sessions:
  • Daily standup reports — Summarize recent PRs, issues, or commits every morning
  • Periodic dependency updates — Check for and apply dependency updates on a weekly basis
  • Recurring data analysis — Generate reports or dashboards from your data at regular intervals
  • Routine code maintenance — Run lint fixes, dead code removal, or test coverage checks on a schedule
  • Monitoring and alerting — Periodically check system health or review logs for anomalies