Developer tools
Give your agents access to PagerDuty.
Agents can look up who is on call, manage incidents, and add notes for responders. Supported incident and service events can start workflows. Access depends on the connection's permissions and your PagerDuty account's features.
5,000 free credits to start.
Agent reach
PagerDuty REST API and incident and service webhooks
The full surface, searched at run time.
Connect with
- API key
- Your own OAuth app
Use a general access key with full access and the email of a PagerDuty user, or your own Scoped OAuth app with the scopes for your workflows, authorized by a user who can create webhook subscriptions. Helios creates the webhook subscription for triggers.
Event triggers
Yes · 8 events
Plus manual, schedule, and signed webhook.
What the agent can do
What agents do with PagerDuty
Helios searches the PagerDuty REST API and incident and service webhooks at run time. The agent finds the endpoint the task needs and calls it.
- Look up and call PagerDuty REST API operations allowed by your connection's permissions.
- Find who is on call for a service, schedule, or escalation policy.
- Create incidents, and acknowledge, reassign, escalate, or resolve them.
- Add notes and status updates to incidents and read their timelines.
- List open incidents by service, team, urgency, or priority.
- Start runs on supported incident and service events, including priority changes, responder requests, and service deletion.
Examples
Tasks you can hand to a PagerDuty agent
Gather context when an incident triggers
The agent reads the incident, checks recent deploys and errors for the service, and posts what it found as a note for the responder.
When a high-urgency incident triggers on the Checkout service, post recent deploys and errors as a note.
- pagerduty·read the incident and its service
- github·list deploys to checkout in the last two hours
- postgres·count failed orders by minute
- pagerduty·add a note with the findings
Note added to incident #1482
- Deploy 3f2a91c went out 14 minutes before the alert
- Failed orders rose from 0.2% to 6% after it
- The same error opened incident #1391 last month
Draft the write-up when an incident resolves
When an incident resolves, the agent reads its timeline and notes and drafts a postmortem for the team to review.
When a P1 incident resolves, draft a postmortem from its timeline and notes.
- pagerduty·read the log entries and notes
- github·find the fix that was merged
- notion·create the postmortem draft
- pagerduty·add a note linking the draft
Postmortem drafted for incident #1482
- Timeline from trigger to resolve: 38 minutes
- Root cause: a config change in deploy 3f2a91c
- Three follow-up actions for review
Answer questions in incident notes
Responders ask the agent a question in a note, and the agent looks it up and answers in another note.
When a note mentions @helios, answer the question in a new note on the same incident.
- pagerduty·read the note and the incident
- postgres·check how many customers are affected
- pagerduty·add a note with the answer
Answered on incident #1482
- 412 customers saw a failed checkout in the last hour
Post who is on call
A scheduled agent reads this week's on-call rotation and posts it to the team's channel.
Every Monday, post who is on call for each service this week to Slack.
- pagerduty·list on-calls for each escalation policy
- code·group by service and level
- slack·post the rotation to #on-call
On-call for this week
- Checkout: Priya, then Sam
- Payments: Alex, then Jordan
Event triggers
Start a run on PagerDuty events
Attach any of these events to a workflow and filter on the fields that matter. Manual, schedule, and signed webhook triggers work too.
| Event | What it means |
|---|---|
incident.triggered | An incident is triggered. Filter by service, team, urgency, or priority. |
incident.acknowledged | A responder picks up an incident. Filter by service, team, urgency, or priority. |
incident.escalated | An incident escalates, so an agent can bring in the next team with context. |
incident.priority_updated | An incident's priority changes. Filter by service, team, urgency, or priority. |
incident.resolved | An incident is resolved. Use its timeline and notes to draft a write-up. |
incident.annotated | Someone adds a note to an incident. Filter on the note's text to answer questions from responders. |
incident.responder.added | Someone is asked to join an incident, so an agent can brief them on what happened so far. |
service.updated | A service changes. Filter by service or team to review changes to critical services. |
Works with
PagerDuty works with the rest of your stack
One agent, one run, several services. It reads from one tool and writes to the next in the same sandbox.
FAQ
Questions, answered
Put an agent on PagerDuty.
Connect it once and describe the task in chat. Then reuse the prompt as a workflow on a schedule or a webhook.