Skip to main content
A trigger is the event that starts a workflow. Every workflow has exactly one, and the trigger you choose decides which conditions you can set and which variables your steps can reference. There are currently over twenty triggers available. They cover incident changes, on-call and alerting events, things people do in Slack, and a recurring schedule to run a workflow like a cron job. Use the trigger reference to find the one you need, and read limits before you build something that needs frequent repetition. To pick a trigger, open Workflows, select New workflow, then choose from the Choose trigger panel. Triggers are grouped there the same way they are on this page.

The Choose trigger panel

​
Trigger reference

API name is the value for the trigger field in the public API and Terraform. Requires is the product or integration a trigger depends on.

​
General

​
User actions

​
Specific changes

​
On-call

​
Other

​
Run a workflow on a recurring schedule

Sometimes you want a workflow to run on a regular cadence, rather than in reaction to a trigger. For example, to send a morning digest of open incidents, or a weekly reminder of who’s on call. This trigger is available to all accounts.

​
Set the cadence

Set the cadence in the trigger card: an interval, an hour, and a timezone.

The cadence controls on a scheduled workflow, set to every day at 09:00

Times are whole hours, read in the timezone you pick, so a 09:00 London workflow follows daylight saving. Runs start within a few minutes of the scheduled time.
Reach out to support or your account manager if you have a use case that requires running a workflow more frequently.

​
What your steps can reference

A scheduled workflow can access a list of active and triage incidents, and a list of incidents in a post-incident status. You can also access any data from your Catalog.

​
Example: a weekday digest of open incidents

1

Choose the trigger and cadence

Create a workflow, choose On a recurring schedule, and set it to every weekday at 09:00 in your team’s timezone.
2

Add a condition (optional)

Leave conditions empty to run every weekday. Add a condition on Active incidents to stay quiet on mornings when nothing is open.
3

Loop over the open incidents

Add a loop over Active incidents, then add Send message to a channel inside it, pointing at your team channel. Reference Each active incident’s name, severity, and lead so each line is useful on its own.
4

Save it live

Save and set live. The first run happens at the next matching occurrence, and you can watch it under Activity on the workflow.
To send one summary message instead of one per incident, skip the loop and reference Active incidents directly in a single Send message to a channel step.

​
Repeating during an incident

To nudge a channel repeatedly while an incident is open, use rule-based suggestions. They prompt responders on a cadence you set, and they adapt to how busy the channel is. For repeat paging when nobody responds, use escalation paths. They have their own levels and repeats, and a workflow can start one with an escalate step. Workflows themselves run from an event or a schedule, and a scheduled workflow runs at most once a day.

​
Test and retrospective incidents

By default a workflow runs on standard incidents only, so a test incident doesn’t set off real automation. Open Advanced settings to add test incidents, retrospective incidents, or both. Workflows never run on declined, canceled, or merged incidents.

​
FAQs

No. Each workflow has exactly one trigger, chosen when you create it and fixed afterwards. To react to two events, build two workflows.
A trigger only appears when your account has what it depends on. On-call triggers need the On-call product, Slack triggers need Slack installed, the debrief trigger needs a calendar integration, and alert workflows need a Pro or Enterprise plan. The Requires column in the trigger reference names each one.
No. Daily, weekdays, weekly, and monthly are the available intervals, and the hour is a whole hour.
Start with An incident is created or changed and narrow it with conditions. Move to a specific trigger such as An incident’s severity changes when your automation is genuinely about that transition, since the specific triggers won’t fire on unrelated edits.