Settings

System-level runtime defaults for analytics and platform behavior

Settings are in read-only mode. Changing the neutral outcome requires neutral-outcome access, and other settings changes require permission-management access.

Outcome Resolution

When multiple rules return different outcomes for the same event, the highest-severity outcome wins.

No outcomes exist yet. Add outcomes on the Outcomes page first.

Neutral outcome

Add outcomes on the Outcomes page before choosing a neutral outcome.

Allowlist rules short-circuit evaluation when they match. They must return {{ neutralOutcomeLabel() }}.

Main rule execution mode

Current default: {{ defaultMainRuleExecutionMode === 'first_match' ? 'Stop on first match' : 'Collect all matches' }}.

The stored neutral outcome is not present in the current outcomes catalog. Select a current outcome and save.

{{ invalidAllowlistRules.length }} allowlist rule{{ invalidAllowlistRules.length === 1 ? '' : 's' }} need attention.

These rules do not currently return the configured neutral outcome {{ neutralOutcomeLabel() }}.

{{ rule.rid }}: {{ rule.error }}

#{{ outcome.severityRank }}

{{ idx === 0 ? 'Highest severity; resolves conflicts first.' : idx === hierarchyOutcomes.length - 1 ? 'Lowest severity fallback.' : 'Lower than outcomes above it.' }}

Top of the list is the most severe outcome. New outcomes are appended to the bottom by default.

Rule Quality Analytics

Default lookback window used by Rule Quality when no explicit lookback is selected.

Lookback Days

Environment fallback: {{ defaultRuleQualityLookbackDays }} days (`EZRULES_RULE_QUALITY_LOOKBACK_DAYS`).

Rule Lifecycle

Control whether edits to already active rules stay live immediately or fall back to the default draft workflow.

Auto-promote edits to active rules

When enabled, saving an edit to an active rule keeps that rule active and updates production immediately. Users still need both modify_rule and promote_rules for this path.

Default fallback: {{ defaultAutoPromoteActiveRuleUpdates ? 'Enabled' : 'Disabled' }}.

AI Rule Authoring

Configure the provider and model used by the AI assistant in the rule editor.

Provider

OpenAI is the only supported provider right now. The select is here so future providers can slot into the same workflow.

Model

Default: {{ aiAuthoringModelOptions[0].value }}. You can keep one of these presets or type any OpenAI model id.

OpenAI API key

Stored key: {{ aiAuthoringSettings?.apiKeyConfigured ? 'configured' : 'not configured' }}. Leave this blank to keep the current key.

Stored API key is configured

Clear it only if you want to remove the current OpenAI key for this organisation.

Enable AI rule authoring

When enabled, the rule editor can request AI-generated rule drafts using the configured OpenAI model.

Curated Rule Quality Pairs

Only configured active outcome→label pairs are used by Rule Quality reports.

Outcome

Label

Outcome	Label	Active	Actions
{{ pair.outcome }}	{{ pair.label }}
No curated pairs configured yet. Add at least one pair to enable Rule Quality reporting.

Strict Mode

Strict mode is an org-level persistence setting. When enabled, served evaluation traffic is treated as the stronger canonical source of truth for downstream analysis, replay, and inspection workflows.

Enable strict mode for this organization

This value is persisted in org-scoped runtime settings and applied independently of any individual operator session.

Change controls

Future enable and disable actions will be reserved for operators with {{ strictModePermissionLabel() }} access.

Disable flow will be intentionally high-friction.

Disabling strict mode requires typed confirmation rather than a casual toggle, and the action is reflected in the dedicated strict-mode audit log.

Default fallback: {{ defaultStrictModeEnabled ? 'Enabled' : 'Disabled' }}.

You can view the current strict mode state, but only operators with {{ strictModePermissionLabel() }} access can change it.

View Strict Mode audit section

Strict mode changes are recorded in the dedicated audit trail.