MetricStream adapter¶

MetricStream is the most-customised of the major GRC platforms. Adapter defaults are best-effort; production deployments almost always provide fieldMappings.

Configuration¶

regulus:
  grc:
    metricstream:
      enabled: true
      base-uri: https://<tenant>.metricstream.com
      auth-token: ${METRICSTREAM_TOKEN}
      intake-app-name: AIControlEvidence    # tenant-specific

Endpoint¶

Default: POST /api/v1/intake/{intakeAppName}/evidence.

MetricStream organises around "applications" (mini-apps) per tenant. Your AI evidence likely targets a tenant-specific app — set intake-app-name accordingly.

Default field mapping¶

GrcEvidenceEnvelope field	MetricStream field
`eventId`	`EvidenceID`
`occurredAt`	`CollectedDate`
`controlFrameworkId`	`Framework`
`controlId`	`ControlID`
`kind`	`EvidenceType`
`actor`	`CollectedBy`
`result`	`Status`
`auditEventLink`	`SourceURL`

For tenants that have custom mini-apps with custom field names, supply a fieldMappings override.

Pre-requisites¶

A mini-app (intake target) defined to receive evidence.
A service-account user with intake permissions.
API access enabled on the tenant.

Caveats¶

Schema drift between mini-apps. A single MetricStream tenant can have many mini-apps; Regulus emits to one. For multi-app deployments, register multiple MetricStreamAdapter beans with distinct intakeAppNames and fieldMappings.