At Polygon, our incident response process is Slack-first. We use incident.io as our primary ChatOps interface for declaring incidents, paging responders, and managing the incident lifecycle — all from within Slack.
incident.io Slash Commands#
incident.io provides the /inc family of Slack commands for incident management. These are the primary commands you'll use during incident response.
/inc (Declare an Incident)#
This is how you declare a new incident. It opens a dialog where you can set the incident name, severity, and other metadata. Once declared, incident.io automatically:
- Creates a dedicated Slack channel for the incident
- Posts an announcement to
#incidents-pos - Begins tracking the incident timeline
/inc update (Post a Status Update)#
Use this to post a status update during an active incident. The update is shared to the incident channel, any configured status pages, and internal subscribers.
/inc close (Resolve the Incident)#
Used to mark the incident as resolved once the IC confirms recovery. This triggers the post-incident flow (schedule a postmortem, etc.).
On-Call and Paging#
Paging an On-Call Group#
To page someone during an incident, use the on-call group Slack handles documented on the Who's On-Call page. For example:
@DevOpsOnCall— pages the DevOps on-call@AppsOnCall— pages the Applications on-call@SecOpsOnCall— pages the SecOps on-call
incident.io's escalation paths handle the routing — Slack first during business hours, phone escalation after hours or if unacknowledged.
Identifying Who's On-Call#
Check the Who's On-Call page or use incident.io's on-call schedules in Slack to see who is currently on-call for each team.
Monitoring#
Datadog#
Our monitoring and alerting is handled by Datadog. Automated alerts flow into #alerts-ps-pos and, when human escalation is needed, into #service-issue-alert. Datadog alerts can also be configured to automatically trigger incident.io incidents based on severity thresholds.
General Principles#
The goal of our ChatOps setup is to ensure that:
- Everything happens in Slack. Responders shouldn't need to context-switch to another tool during an active incident.
- Paging is fast and easy. Anyone can escalate to the right on-call group with a single Slack @-mention.
- The timeline is automatic. incident.io tracks key events (declaration, status updates, resolution) so the First Responder can focus on capturing context rather than timestamps.
- Triggering incident response has a low barrier. Anyone can declare an incident using
/inc. See severity levels for guidance on when to declare.