Most organizations managing major incidents realized that every moment matters. Context-switching between different tools – with multiple web and chat surfaces having to be open to collect information causes friction. The cost of context-switching during an outage is even more painful. These are the problems that PagerDuty’s Slack Transformation just closed. 

With this in mind, PagerDuty has rebuilt its Slack integration into something closer to a command center: agentic AI teammates, incident cards you design yourself, custom fields pulled straight from your data sources, and full incident lifecycle — declare, mobilize, diagnose, update, resolve, review — that runs almost entirely without leaving a dedicated channel. PagerDuty isn’t a notifications app with a Slack plugin anymore. It’s an operations platform that happens to use chat as a surface.

And today, we highlight how PagerDuty’s SRE Agent “Paige” can change the game in how context is provided and how relevant information can be used to diagnose issues and provide steps towards auto-remediation.

Paige: Your Incident Commander and Agentic Teammate

The biggest shift is who’s in the channel with you. Paige, PagerDuty’s SRE Agent, joins the incident as something closer to an incident commander than a chatbot. She has access to connectors, tools, workflows, and skills that let her pull context from your system history in PagerDuty, triage and diagnose what’s going wrong, and push toward auto-remediation — not just surface a dashboard link and step back.

A few ways that Paige shows up in the channel:

Automated workflows and remediation. Paige can enact a workflow and complete steps automatically. As she pulls context from connected systems, she can trigger the workflows and automations tied to an incident type — running diagnostic checks, kicking off remediation actions, or escalating based on what she finds — turning what used to be a responder manually working a runbook into a set of steps Paige is already executing in parallel.

AI-generated status updates. Keeping stakeholders current during an incident is its own job, and it usually falls on whoever’s already heads-down on the fix. Paige takes that off responders’ plates: she synthesizes what’s happened so far — diagnosis, actions taken, current state — into a clear status update, so /pd status posts something a stakeholder can actually absorb in the Announcement Channel, without pulling a responder out of the Dedicated Channel to write it.

Alongside Paige, the Scribe Agent joins the incident bridge, transcribes the call, and posts a context-aware summary straight into the Slack channel — turning the scramble to reconstruct “what happened and when” into a one-click starting point for the post-incident review. 

In addition to announcing our SRE Agent’s new brand, I’d like to highlight other customizations that make working in Slack much easier.

Custom Cards: Built for the Right Audience

For years, the incident card Slack posted was static — every team saw the same fixed fields (status, priority, assignee) whether or not those were the fields that mattered to them. Important context routinely went missing, and responders left Slack mid-incident to hunt it down on pd.com. For an SRE heads-down on a data center outage, that’s exactly the friction Paige and Custom Cards together are designed to remove.

Custom Cards make the incident card configurable — critically, configurable per channel type. Because Notification, Dedicated, and Announcement Channels serve different audiences, each gets its own card template instead of one design forced across all three. The card a responder sees in a Dedicated Channel can look nothing like the card a stakeholder sees in an Announcement Channel, even on the same incident.

You configure templates in Integrations > Slack Configuration inside PagerDuty: pick the channel type (Notification, Dedicated, or Announcement) and open the Card Builder for it. From there, admins can show or hide the default system fields (status, priority, assignee, and similar) and add custom fields already set up in the web app for that incident type. Save it, and the template applies to every incident card posted in that channel type going forward — no per-incident setup.

Custom Fields, surfaced where the work happens

Card Builder is only as useful as the fields available to put in it — that’s what Custom Fields solve. Fields configured in the PagerDuty web app become selectable inside the Slack Card Builder for the relevant channel type. Card Builder also filters by incident type, so building a template for a network outage only shows you the custom fields actually configured for network outages.

See our documentation to learn more.

Autonomous operations: End-to-End Incident Management

Put Paige, channel types, and custom cards together, and the incident lifecycle stays inside Slack start to finish:

Multiple agents, channel types, and custom cards together, and you get an incident lifecycle that stays inside Slack from start to finish. Responders can declare an incident with /pd declare, and PagerDuty’s default mobilization routing — paired with automatically created Dedicated Channels for top-priority incidents — gets the right people into the right room fast, without waiting on someone to manually spin up a channel. From there, the SRE Agent helps diagnose what’s happening and provides steps towards remediation, /pd status keeps stakeholders current in the Announcement Channel while responders keep working the Dedicated Channel, and /pd resolve closes it out. The Scribe Agent’s transcript, sitting in the channel the whole time, becomes the seed for a fast, low-friction post-incident review.

For an SRE managing critical data center infrastructure, that’s the real story here. PagerDuty in Slack isn’t a notification feed with slash commands anymore — it’s a configurable, increasingly autonomous operations command center, and the chat window is just the interface.

Watch the full demo experience: