Database Workflow Skill

Overview

This skill covers common GitHub Actions patterns for running Liquibase-driven database schema updates.

Dojo360 provides a reusable database.yml workflow. You typically wire it into your repo as a workflow that supplies:

• Team metadata inputs (aide-id, team-name, domain, cloud-type, environment)
• Liquibase parameters (changelog file, command, contexts)
• Database connection details (via GitHub Secrets, PRM/Volcan, or platform-native secret stores)

Workflow Reference

Repository: dojo360/pipelines-workflows
Workflow: .github/workflows/database.yml
Docs: https://github.com/dojo360/pipelines-workflows/tree/main/web/database/index.md

Key Features

• Liquibase-based schema management (update/validate/rollback patterns)
• Schema versioning via changelog files
• Rollback support (tag/count based patterns)
• Multi-database support (varies by platform/driver)

Prerequisites

• A Liquibase changelog exists in the repo (example paths: db/changelog.xml, liquibase/changelog.yaml)
• Database connection details are available to the workflow (avoid hardcoding; prefer OIDC + vault-backed secrets)
• GitHub Environments exist if you rely on environment-level approvals/protection rules

Required Inputs (Typical)

Exact inputs can vary by Dojo360 version; confirm against the Dojo360 docs.

Input	Type	Description
aide-id	string	AIDE ID used to fetch team metadata
team-name	string	Team name for metadata lookup
domain	string	Domain for metadata lookup
cloud-type	string	Cloud provider (e.g., awsOptum, azureOptum, gcp)
environment	string	Target environment (e.g., dev, qa, cert, prod)

Database / Liquibase Inputs (Typical)

Input	Type	Description
liquibase-command	string	Liquibase command (e.g., update, validate, rollback, updateSQL)
liquibase-changelog-file	string	Path to changelog file in repo
liquibase-contexts	string	Optional contexts (comma-separated)

Templates

Ready-to-copy examples are in .github/skills/github-workflows-dojo360-database/templates/:

• basic-database-management.yml
• database-management-with-rollback.yml
• database-management-pr-validation.yml

Notes / Gotchas

• Never commit DB credentials or JDBC URLs into workflow YAML; use GitHub Secrets or platform secret stores.
• Use GitHub Environments to enforce approvals for higher environments.
• When possible, run validate in PRs and reserve update for protected branches.