Skip to main content
The Incidents page (/incidents) is your full list view of every incident — open, closed, and in-progress — across your environment. You can search, filter, and drill into any incident from here.

What is an incident?

An incident is opened automatically when an alert fires and sureops determines it meets the threshold for agent-driven response. Each incident tracks the full lifecycle from first detection through root-cause diagnosis, fix execution, metric verification, and formal closure. Incidents are scoped to an environment. Use the context switcher in the sidebar to view incidents for a different environment.

Lifecycle stages

Every incident progresses through up to six stages:

Severity levels

The severity is set by the agent during the triage stage based on the alert payload, error rate, affected service tier, and blast radius. Severity is only settable at incident-creation time (for manually created incidents) or by the agent during triage — it’s read-only on the Incident Hub, with no severity-edit control there.

Status badges

The badge set differs depending on which view you’re looking at:
  • Active/All tab (card grid view) — each card shows an environment badge plus either a severity badge (open incidents) or a closure-reason badge (closed incidents shown here) — not both together — and a scope badge below.
  • Resolved tab (list view) — each row shows a severity badge, a closure-reason badge, and a scope badge together — but no environment badge.
Once an incident is closed, its closure-reason badge describes how it ended:
  • Resolved by fix — an approved remediation was applied and verified
  • Recovered on its own — the incident self-healed before a fix was needed
  • Closed after escalation — the incident was escalated to a human and closed outside the automated flow
  • Superseded — closed because it was folded into another incident
  • No remaining alerts — closed because the triggering alerts cleared on their own
While an incident is still open, you may also see a flapping-related badge when the agent has paused because it’s treating the pattern as flapping or self-healing.

Filtering incidents

Use the filter bar to narrow the list:
  • Status — a three-way tab: All, Active, or Resolved
  • Severity — P1 through P4
  • Search — matches on incident title or external ID, over the incidents currently loaded on the page (not a full-text or service-name search)
There is no separate stage filter on this page — use the Status tabs and Severity filter together, or open an individual incident to see its current stage.
Status and severity filters persist in the URL, so you can bookmark or share a filtered view. Search is local to your current session and does not persist in the URL or survive a page refresh.

Creating an incident manually

In most cases incidents are opened automatically by incoming alerts. You can also open an incident manually:
  1. Click New Incident in the top-right of the Incidents page.
  2. Enter a Title.
  3. Select the Severity and Environment.
  4. Optionally add a Description.
  5. Click Create Incident — the Incident Commander agent picks it up immediately.
Manually created incidents follow the same lifecycle as alert-triggered ones. The difference is that the detection stage context will be minimal until the agent queries telemetry.
Click any row in the list to open the Incident Hub for that incident. The hub shows the full stage-by-stage detail view, timeline, agent analysis, and HITL approval controls.

Incident to problem relationship

sureops automatically promotes a closed incident to a problem record if it exhibits recurrence patterns or a root cause that warrants structural follow-up. You can also manually promote an incident to a problem at any time — including a still-open one — using the Promote to problem button on its Incident Hub; it doesn’t require the incident to be closed first. The incident list itself doesn’t show a problem-record link — that link lives on the Incident Hub header for the incident once it’s linked to a problem.