The @Rootly Agent is now available in Slack. Mention @Rootly during an incident and it reads your Rootly incident data, the channel discussion, and bridge call transcripts (Rootly Meeting Scribe) in real time. It answers questions, takes actions, and reports back in the thread, so everyone stays up to speed without leaving the channel.
Why it matters
During an active incident, the response work is only half the job. A responder joins twenty minutes late and has to scroll back to catch up. A comms lead needs a customer update while the commander is still triaging. Leadership wants status without interrupting the people fixing the problem. Each of these is a context switch: a question dropped in the channel, a jump into the Rootly web app, a DM that breaks someone's focus.
The @Rootly Agent lives in the channel where the incident is already running. Instead of switching tools to find who's on call, change a severity, or draft an update, you ask in plain language and it does the work. It's grounded in the real incident: your Rootly data (services, on-call, roles, action items) plus the discussion in the channel and on the bridge call (Rootly Meeting Scribe) where decisions get made. Instead of a generic chatbot, you get a responder that's caught up on what's happened, what's been tried, and what's been decided…and ready to act on it.
How it works
Mention @Rootly in an incident channel and ask. A few examples you can try below:
The things teams reach for most:
- Catch up. "Catch me up" returns a structured brief: current status, the leading theory, affected services, open action items, and who's involved.
- Find who's on call. "Who's on call for payments?" answers across teams, services, schedules, and escalation policies, including who was on call when the incident started.
- Catch what's slipping. "Did I miss any action items? Assign them to me." It cross-references the channel discussion against what's already tracked, surfaces commitments that were said but never logged, and assigns them to you.
- Take action. Change severity, status, title, roles, and custom fields; create and manage action items, page through Rootly On-Call, and draft or publish a status-page update.
- Ask for a recommendation. "I'm not sure what severity this should be: set it for me." It weighs the situation against your team's own severity definitions, recommends a level, and makes the change.
- Learn from the past. "Find similar incidents" surfaces related incidents so your team isn't solving a problem from scratch.
Every action runs as you, under your existing Rootly permissions. The @Rootly Agent can't do anything in Slack you couldn't already do in the web app, and each change is attributed to you in the audit trail. You can also talk to the Rootly Agent in the Slack side-pane or message it directly if you're mindful of adding noise to the incident channel.
Getting started
Rootly Workspace Admins can enable the @Rootly Agent in Slack from Configuration → Rootly AI → Enable @Rootly in Slack. After toggling it on, a Slack workspace Admin reconnects the Slack integration so the new permissions take effect.
See the documentation for setup steps and the full list of what you can ask.
.webp)
