Configure Severity and Priority Labels
Configure severity levels for incidents and priority levels for alerts to match your organization's terminology and classification system.
Overview
By default, Harness AI SRE uses standard severity and priority labels. You can configure these labels organization-wide to:
- Match your existing incident management terminology
- Provide clearer context for your team
- Align with your organization's severity classification standards
- Improve consistency across incident response workflows
Important: Custom labels only change the display text. The underlying severity/priority values remain the same, ensuring compatibility with integrations, APIs, and historical data.
Access Custom Options
- Navigate to Organization Settings
- Under Settings mainly used in AI SRE, click Severities & Statuses for AI SRE
- Use the tabs to configure:
- Incidents tab - Configure severity levels
- Alerts tab - Configure priority levels
Configure Incident Severity Levels
Default Severity Labels
| Level | ID | Default Label | Description |
|---|---|---|---|
| SEV0 | "0" | SEV0: Critical | Highest severity |
| SEV1 | "1" | SEV1: Major | High severity |
| SEV2 | "2" | SEV2: Moderate | Moderate severity |
| SEV3 | "3" | SEV3: Minor | Low severity |
| SEV4 | "4" | SEV4: Cosmetic | Lowest severity |
Configure Severity Labels
To configure severity labels:
- Go to Organization Settings → Severities & Statuses for AI SRE
- Select the Incidents tab
- Click the Edit icon next to any severity level
- Update the:
- Label - The display name (max 20 characters)
- Description - Additional context for your team
- Click Save
Example customizations:
- Change "SEV0: Critical" to "P0: Business Critical"
- Change "SEV1: Major" to "P1: Production Down"
- Change "SEV2: Moderate" to "P2: Degraded Service"
Using Custom Severity in Workflows
Custom labels appear throughout AI SRE:
- Incident cards and detail pages
- Filters and search
- Status updates
- Reports and analytics
- Slack notifications
- Email notifications
API and Integration Usage:
When using the API or configuring runbook triggers, always use the original ID values ("0", "1", "2", "3", "4"), not the custom labels.
# Runbook trigger example
trigger:
condition:
field: severity
operator: equals
value: "0" # Use ID, not custom label
Go to Incident Fields for more information about using severity values in integrations.
Configure Alert Priority Levels
Default Priority Labels
| Level | ID | Default Label | Description |
|---|---|---|---|
| P1 | p1_critical | P1: Critical | Critical priority |
| P2 | p2_error | P2: Error | High priority |
| P3 | p3_warning | P3: Warning | Medium priority |
| P4 | p4_info | P4: Info | Low priority |
Configure Priority Labels
To configure priority labels:
- Go to Organization Settings → Severities & Statuses for AI SRE
- Select the Alerts tab
- Click the Edit icon next to any priority level
- Update the:
- Label - The display name (max 20 characters)
- Description - Additional context for your team
- Click Save
Example customizations:
- Change "P1: Critical" to "Urgent"
- Change "P2: Error" to "High"
- Change "P3: Warning" to "Medium"
- Change "P4: Info" to "Low"
Using Custom Priority in Workflows
Custom labels appear throughout AI SRE:
- Alert cards and detail pages
- Alert rules
- Filters and search
- Reports and analytics
- Slack notifications
- Email notifications
API and Integration Usage:
When using the API or configuring alert rules, always use the original ID values (p1_critical, p2_error, p3_warning, p4_info), not the custom labels.
# Alert rule example
alert_rule:
condition:
priority: p1_critical # Use ID, not custom label
action:
create_incident: true
Best Practices
Label Design
- Keep labels concise - Maximum 20 characters
- Use consistent naming - Follow a standard pattern (e.g., "P0: Business Critical", "P1: Production Down")
- Avoid abbreviations - Unless universally understood in your organization
- Include severity level - Help users understand the hierarchy at a glance
Descriptions
- Add context - Explain when to use each severity/priority level
- Define impact - Describe the typical impact for each level
- Provide examples - Include example scenarios
- Update regularly - Keep descriptions current as processes evolve
Team Alignment
- Document changes - Announce custom labels to your team
- Train responders - Ensure everyone understands the new terminology
- Update runbooks - Reference custom labels in response procedures
- Gather feedback - Periodically review effectiveness with your team
API and Integration Considerations
- Test integrations - Verify external systems work with custom labels
- Document mappings - Create a reference showing ID to label mappings
- Update documentation - Update internal runbooks and procedures
- Monitor for issues - Check that custom labels don't break workflows
Important Notes
What Changes
- Display labels throughout the UI
- Labels in notifications (Slack, email)
- Labels in exports and reports
- Labels in search and filters
What Doesn't Change
- Underlying ID values
- API request/response formats
- Integration mappings
- Historical data
- Database storage format
Organization-Wide Impact
Custom severity and priority labels apply to:
- All projects in the organization
- All teams and users
- All incidents and alerts (past and future)
- All integrations and workflows
Troubleshooting
Custom labels not appearing
- Refresh the page after saving changes
- Check browser cache (try incognito/private mode)
- Verify you have organization admin permissions
Integrations sending incorrect severity
- Integrations send original severity values
- Custom labels are applied on display only
- Check integration mappings in Incident Fields
API requests failing
- Always use ID values in API calls, never custom labels
- See Incident Fields for correct ID values
Example Configurations
Enterprise Standard
SEV0 → "P0: Business Critical"
SEV1 → "P1: Major Incident"
SEV2 → "P2: Significant Issue"
SEV3 → "P3: Minor Issue"
SEV4 → "P4: Informational"
SRE Team
SEV0 → "Critical: All hands"
SEV1 → "High: Oncall + Manager"
SEV2 → "Medium: Oncall"
SEV3 → "Low: Best effort"
SEV4 → "Info: Track only"
Customer Impact Based
SEV0 → "Total Outage"
SEV1 → "Major Degradation"
SEV2 → "Partial Impact"
SEV3 → "Minimal Impact"
SEV4 → "No Customer Impact"