PagerDuty

Automate on-call incident response. Create and run the full incident lifecycle, fire alerts and change events through the Events API v2, and administer services, escalation policies, schedules, users, teams, and maintenance windows.

1 trigger 68 actions available

On New Triggered Incident fires when a service pages

Agent reads the incident title, urgency, and the alert payload that opened it

Agent correlates the incident against recent deploys and open maintenance windows

Agent posts the incident summary and a runbook link to the on-call channel

Agent pulls who is on call now and names the responder in the summary

The on-call engineer acknowledges the incident in PagerDuty before the agent runs any recovery step

What This Integration Enables

FlowRunner agents use PagerDuty to turn a raw signal into a well-formed incident and to keep on-call operations in sync with the rest of the stack. An agent can open an incident when a check fails, post a status update to subscribers, snooze or re-assign, request an extra responder, and record a change event so the incident correlates with the deploy that caused it. It can keep schedules, escalation policies, and team membership synchronized from other systems, and look up who is on call right now to route the page correctly. What it does not do is run the response on its own. The engineer who owns the page owns the decisions.

Without FlowRunner

Noisy first minutes The responder opens several tools to figure out what paged and why

No correlation Nobody links the incident to the deploy that likely caused it

Manual context gathering On-call status, recent changes, and runbook links are collected by hand

With FlowRunner

Context on arrival The agent posts urgency, likely cause, and the responder name with the page

Change correlation Recent deploys and change events are already attached to the incident

Human acknowledges first Recovery steps wait for the on-call engineer to acknowledge and decide

Use Case Scenarios

Well-formed incidents from monitoring signals

A monitoring export lands with a set of failing checks. The agent reads each row and calls Trigger Alert into the Events API v2, reusing a deduplication key so a recovered check clears itself later with Resolve Alert. For the checks that represent a real outage, the agent calls Create Incident on the right service, attaches the failing check details, and correlates recent deploys with Send Change Event. The on-call engineer opens a fully described incident instead of a bare alert.

Deploy-aware maintenance windows

A deployment workflow is about to ship. Before it does, the agent calls Create Maintenance Window on the affected services to suppress the expected alert noise, records the deploy with Send Change Event, and removes the window when the deploy is confirmed healthy. Expected noise stays quiet. Real problems still page.

Incident acknowledgement gate before recovery

An incident triggers on a service that has an automated recovery step. The agent posts the incident and the proposed recovery action to the on-call channel, then waits. It does not run the recovery. Only after the on-call engineer acknowledges the incident in PagerDuty does the agent proceed, and it records the acknowledgement as part of the incident timeline.

Human-in-Loop Highlight

The human-in-loop moment here is the acknowledgement gate. When an incident triggers and a recovery action is on the table, the agent stops at exactly the point where judgment matters. It opens the incident, posts the context and the proposed action, and holds. The agent watches for the on-call engineer to acknowledge the incident in PagerDuty, using List Log Entries to confirm the acknowledgement is real and to capture who did it. Only then does it run the recovery step. An unacknowledged page never triggers an automated action against production. The responder decides the response, and the timeline shows they did.

Agent processes routinely

Detects exception requiring judgment

Clear match Continues automatically

Ambiguous Routes to human via preferred channel

Human decides

Agent resumes with decision

Agent Capabilities

68 actions

Incidents

Create Incident Open a new incident on a service, paging the on-call responders. Used to raise an alert from a flow when a check fails or a customer reports an outage.
List Incidents List incidents, filtered by status, service, or team. Used to build a live incident view or a digest.
Get Incident Retrieve a single incident's full detail.
Update Incident Change an incident's status, urgency, assignment, or priority.
Merge Incidents Merge duplicate incidents into one. Used to collapse a storm of related pages into a single response.
Snooze Incident Snooze an incident for a set period.
Create Note Add a note to an incident. Used to record context or a decision on the incident timeline.
List Notes List the notes on an incident.
Create Status Update Post a status update to an incident's subscribers.
Create Responder Request Request that an additional responder be paged into an active incident. Used to pull in extra help during a live incident.
List Alerts List the alerts that belong to an incident.
Get Alert Retrieve a single alert on an incident.
Update Alert Update an alert on an incident, such as resolving it.
List Log Entries List the log entries for an incident. Used to pull the incident's audit timeline and confirm an acknowledgement.

Services

List Services List the services in the account.
Get Service Retrieve a single service.
Create Service Create a service. Used to onboard a new system into on-call coverage.
Update Service Update a service's configuration.
Delete Service Delete a service.

Escalation Policies

List Escalation Policies List the escalation policies in the account.
Get Escalation Policy Retrieve a single escalation policy.
Create Escalation Policy Create an escalation policy.
Update Escalation Policy Update an escalation policy.
Delete Escalation Policy Delete an escalation policy.

Schedules

List Schedules List the on-call schedules.
Get Schedule Retrieve a single schedule.
Create Schedule Create an on-call schedule.
Update Schedule Update a schedule.
Delete Schedule Delete a schedule.
List Overrides List the overrides on a schedule.
Create Override Create a schedule override. Used to cover an on-call gap.
List On-Calls List who is on call now across schedules and escalation policies. Used to name the right responder during a live incident.

Users

List Users List the users in the account.
Get User Retrieve a single user.
Create User Create a user.
Update User Update a user.
Delete User Delete a user.
List Contact Methods List a user's contact methods.
List Notification Rules List a user's notification rules.

Teams

List Teams List the teams in the account.
Get Team Retrieve a single team.
Create Team Create a team.
Update Team Update a team.
Delete Team Delete a team.
Add User Add a user to a team.
Remove User Remove a user from a team.
Add Escalation Policy Attach an escalation policy to a team.
Remove Escalation Policy Detach an escalation policy from a team.

On-Call

List Priorities List the incident priority levels configured in the account. Used to set priority on a new incident.

Maintenance Windows

List Maintenance Windows List the maintenance windows.
Get Maintenance Window Retrieve a single maintenance window.
Create Maintenance Window Create a maintenance window to suppress expected alerts around a deploy.
Update Maintenance Window Update a maintenance window.
Delete Maintenance Window Delete a maintenance window.

Business Services

List Business Services List the business services.
Get Business Service Retrieve a single business service.
Create Business Service Create a business service.
Update Business Service Update a business service.
Delete Business Service Delete a business service.

Events

Trigger Alert Fire an alert into the Events API v2 with a routing key and deduplication key. Used to open or update an alert from a monitoring signal.
Acknowledge Alert Acknowledge an Events API v2 alert by deduplication key.
Resolve Alert Resolve an Events API v2 alert by deduplication key, so recoveries clear automatically.
Send Change Event Record a change event, such as a deploy, in the Events API v2 so incidents correlate with changes.

Triggers

1 triggers

Event Triggers

On New Triggered Incident Fires when a new incident is triggered in PagerDuty. Starts an incident-response workflow the moment a service pages, before a human has acknowledged, so the agent can gather context and post it to the on-call channel while the responder is still reading the page.

Start building with PagerDuty

$100 in credits. No card required. Connect in minutes.

Start Building Free Talk to us