What problem does it solve? Managing MariaDB on Kubernetes requires authoring many interdependent custom resources across 12 CRDs, and mistakes in Galera topology, backup retention, or image format versions cause outages or data loss. This Skill provides complete, version-accurate CRD references and working manifests so agents generate correct MariaDB operator resources on the first attempt. ## Core Features & Use Cases - Full CRD coverage: Documents all 12 CRDs under k8s.mariadb.com/v1alpha1 — MariaDB, MaxScale, Backup, PhysicalBackup, Restore, PointInTimeRecovery, Database, User, Grant, Connection, SqlJob, and ExternalMariaDB — with field tables and YAML examples. - HA and data protection patterns: Provides ready manifests for Galera clusters with automatic failover, MaxScale read/write splitting, scheduled S3/Azure backups, and point-in-time recovery to a target timestamp. - Version-aware guidance: Flags the v26.3.0 image format change (image.repository + image.tag) and v26.6.0 features like password rotation and cross-cluster replication. - Use Case: Ask the agent to deploy a 3-node Galera MariaDB cluster with hourly S3 backups, a MaxScale proxy, and an application database with user and grants — it produces the full multi-document YAML. ## Quick Start Ask the agent to create a MariaDB Galera cluster with three replicas, a scheduled S3 backup, and a Database, User, and Grant for your application.
