Skip to content
Nebula Logger

Tags

Where a scenario applies to the whole transaction, a tag applies to a specific entry. Multiple tags per entry are supported, tag names are free-text, and new tag names auto-create new LoggerTag__c records.

Adding tags

Chain .addTag(...) off the level-specific call.

Logger
  .warn('Credit check returned warnings')
  .setRecord(account)
  .addTag('credit-check')
  .addTag('customer-onboarding');


Logger.saveLog();

In LWC:

this.logger.info('Payment submitted').addTag('payment').addTag('checkout').setRecord(this.paymentId);


await this.logger.saveLog();

In Flow, the Tags input on any Add Log Entry* action accepts a semicolon-delimited list: credit-check;customer-onboarding.

In OmniStudio, pass tags as a List<String> input to a newEntry Remote Action call.

Storage model

  • Each unique tag name becomes a LoggerTag__c record. Reusing an existing tag name reuses the record.
  • The join between a LogEntry__c and its tags is stored in LogEntryTag__c (a junction object).

The console renders tags on the log entry record page and lets you filter list views by them.

Choosing a tagging mode (unlocked package only)

The default storage above (LoggerTag__c + LogEntryTag__c) is one of two supported modes. The unlocked package also supports storing tags as Salesforce Chatter Topics on each LogEntry__c record via the standard Topic and TopicAssignment objects.

AspectDefault: LoggerTag__c + LogEntryTag__cAlternative: Topic + TopicAssignment
AvailabilityBoth packagesUnlocked package only
Data locationCustom objectsStandard Chatter objects
Data visibilitySharing/OWD on LoggerTag__c. Junction records inherit access from the related entry + tag.Every Chatter user can see all Topic records org-wide. TopicAssignment records are gated by access to the entry.
Reports & dashboardsFull custom-object reporting.Chatter Topics have partial reporting support.
List view filteringStandard LoggerTag__c filters.Chatter Topics filter list views natively.
Best forMost orgs. Full control, private-by-default, reportable.Orgs already invested in Chatter Topics and comfortable with tag names being globally visible.

Choose the mode that matches your access-control posture. Neither mode is available for admins to switch on the fly - the choice is a configuration decision made once per org.

Taxonomy conventions

Because tag names are free-text, teams need a convention or the taxonomy sprawls. Adopt a prefix-per-namespace pattern early.

Common namespaces:

  • domain:* - what part of the business (domain:billing, domain:orders).
  • feature:* - what feature or capability (feature:credit-check, feature:sync).
  • incident:* - live incidents (incident:2026-08-12-payments).
  • severity:* - if you need more nuance than the logging level (severity:actionable, severity:info-only).

Prefixing keeps the taxonomy scannable and lets a list view filter cover a whole namespace at once.

Auto-tagging via LogEntryTagRule__mdt

For tags that should attach to entries automatically based on their content, define a LogEntryTagRule__mdt record. Each rule matches on a field’s value (Message__c, OriginLocation__c, etc.) and applies one or more tags when it matches.

Useful when a tag should be applied consistently across every entry meeting a condition - e.g. auto-tag anything logged from PaymentService.processPayment with feature:payments, or auto-tag anything whose message contains Rate limit with incident:rate-limiting.

LogEntryTagRule example

Auto-applied tags stack with any tags added at the point of logging - they don’t replace them.

Tags vs scenarios vs logging level

Three orthogonal axes.

  • Logging level is a severity gradient. ERROR -> FINEST. Determines whether an entry persists at all (via LoggerSettings__c.LoggingLevel__c).
  • Scenario is the business process the whole transaction is part of. One per transaction. Drives per-scenario retention and reporting.
  • Tags are freeform labels on individual entries. Many per entry. Drives taxonomy filtering.

Don’t overload one axis to do another’s job. If you find yourself using tags to indicate severity, use logging level instead. If you find yourself using scenarios to label individual entries, use tags instead.

Where next