Sev-1
Requires major incident command and leadership visibility
Shared operating model for how SysOps classifies service impact, assigns command roles, controls communication cadence, and governs AI involvement across severity levels.
The severity framework is now route-native in `apps/web`, giving the platform one shared place to declare impact, cadence, roles, and AI boundaries under pressure.
Requires major incident command and leadership visibility
Requires service owner control and structured update cadence
Handled by operational leads with visible customer impact tracking
Low-impact issue or request handled in normal queue flow
Severity should reflect business consequence, not just technical noise or who is asking the loudest.
| Severity | Business Impact | Operational Pattern | Required Roles | Approved AI Involvement |
|---|---|---|---|---|
| Sev-1 | Major service outage, externally visible disruption, or cross-division impact that materially impairs agency operations. | Major incident command activated immediately with explicit checkpoint cadence and leadership visibility. | Incident commander, technical lead, communications lead, business liaison, executive sponsor when needed. | Tier 4 specialist support only for evidence gathering, timeline drafting, communication drafts, and vendor packet preparation. |
| Sev-2 | Shared service degraded for multiple users, critical workflow impaired, or substantial risk to business timing. | Dedicated incident lead, rapid triage, structured updates, and possible vendor or cross-team escalation. | Service owner, incident lead, technical resolver, communications approver as needed. | Tier 3 or Tier 4 support for diagnostics synthesis, workaround analysis, and draft communications under human approval. |
| Sev-3 | Single service impaired for a limited user group, workaround exists, or business impact is meaningful but contained. | Managed through normal operations with active owner oversight and visible aging. | Resolver owner, team lead, service owner when aging or recurrence increases. | Tier 2 or Tier 3 support for triage notes, known-error retrieval, duplicate analysis, and next-step drafting. |
| Sev-4 | Routine issue, limited inconvenience, or low-risk request with no broad service consequence. | Normal queue handling and standard prioritization. | Assigned analyst or service desk owner. | Tier 1 or Tier 2 support for intake clarification, knowledge retrieval, checklist preparation, and status drafting. |