Schedules

Schedules enable agents to run automatically at specified times. Whether it is a daily report, hourly sync, or weekly cleanup, schedules handle the timing so you do not have to.

Why Schedules?

Automate recurring work without manual intervention:

• Daily reports -- Generate summaries every morning
• Data sync -- Keep systems in sync hourly
• Maintenance -- Weekly cleanup and optimization
• Monitoring -- Check system health every 5 minutes

Frequency Options

Frequency	Description	Additional Config
Once	Run once at a specific time	Run-at timestamp
Hourly	Every hour	Minute offset
Daily	Every day	Hour and minute
Weekly	Every week	Day of week, hour, minute
Monthly	Every month	Day of month, hour, minute
Cron	Custom cron expression	Cron string

Cron Expression Reference

For complex schedules, use standard cron syntax with five fields: minute, hour, day of month, month, and day of week.

Pattern	Description
Every minute	Runs once per minute
Every hour (on the hour)	Top of each hour
Daily at midnight	Once per day
Weekdays at 9 AM	Monday through Friday
Every 15 minutes	Four times per hour
First of every month	Monthly on the 1st
Every Sunday	Weekly on Sunday

Timezone Matters

Always specify a timezone when creating schedules. Without one, schedules default to UTC, which can lead to surprising run times for your team.

Creating Schedules

From the Dashboard

1. Go to Agents and select an agent, then navigate to Schedules
2. Click + Create Schedule
3. Choose a frequency (preset or custom cron)
4. Set the timezone
5. Configure retry and timeout settings
6. Add input parameters for the agent
7. Click Create

Schedule Configuration Options

Setting	Description	Default
Frequency	How often the schedule runs	Required
Timezone	IANA timezone for execution	UTC
Max Retries	Retry count on failure	0
Timeout	Maximum execution time per run	No limit
Start Date	When the schedule becomes active	Immediately
End Date	When the schedule expires	Never
Enable Memory	Agent remembers state between runs	Off

Date Bounds

You can restrict a schedule to a specific date range. For example, a "Q4 Campaign" schedule can start on October 1st and automatically stop on December 31st.

Memory Persistence

Enable memory persistence for incremental processing. The agent remembers what it processed in previous runs and can pick up where it left off -- essential for data sync and batch processing scenarios.

Managing Schedules

From the dashboard or via API, you can:

• Pause -- Temporarily stop a schedule without deleting it
• Resume -- Reactivate a paused schedule
• Run Now -- Trigger an immediate run outside the normal timing
• Update -- Change frequency, timezone, or input parameters
• Delete -- Permanently remove a schedule

Monitoring Schedules

Each schedule tracks:

• Last run -- When it last executed and whether it succeeded
• Next run -- The next scheduled execution time
• Run count -- Total executions
• Failure count -- How many runs have failed
• Next 5 runs -- Preview of upcoming execution times

You can also view the full run history with status, errors, and timing for each execution.

Best Practices

Stagger Heavy Schedules

Avoid running all resource-intensive tasks at the same time. Spread them across different hours to balance load.

Configure Retries for Network Operations

External API calls may fail transiently. Setting 2-3 retries handles most intermittent issues automatically.

Use Memory for Incremental Processing

Enable memory persistence so agents can track what was already processed and avoid redundant work on each run.

Set Reasonable Timeouts

A schedule without a timeout can run indefinitely if something goes wrong. Match the timeout to the expected duration of the task.

---

Next: Learn about Triggers for webhook-based agent execution.