Help us improve
Share bugs, ideas, or general feedback.
From data-architecture
Manage database schema changes while maintaining backwards compatibility. Handle migrations, versioning, and zero-downtime deployments. Use when evolving data models in production systems.
npx claudepluginhub sethdford/claude-skills --plugin architect-data-architectureHow this skill is triggered — by the user, by Claude, or both
Slash command
/data-architecture:schema-evolutionThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Evolve database schemas safely while maintaining backwards compatibility and minimizing downtime.
Safe schema migrations, backward compatibility, zero-downtime deployments, and rollback strategies.
Creates, validates, executes, and rolls back schema migrations for PostgreSQL, MySQL, MongoDB using Flyway, Alembic, Prisma, Knex.
Build zero-downtime database migrations — forward SQL, rollback SQL, deployment sequence. Use when asked to "write migration", "schema change", "add column", "rename table", "drop column", or "migrate safely".
Share bugs, ideas, or general feedback.
Evolve database schemas safely while maintaining backwards compatibility and minimizing downtime.
You are planning schema changes in production systems. Design migrations that don't break deployed clients or services. Read current schemas, deployment frequency, and rollback requirements.
Based on production database migration best practices:
Plan Phase 1 (Additive): Add new columns/tables. Code ignores new columns for now. Deploy database changes. All existing code still works.
Plan Phase 2 (Dual Write): Update application code to write to both old and new schema. Reads can use either. Let data double-write for period (hours to days).
Plan Phase 3 (Migrate Data): Background job migrates existing data to new schema. Backfill new fields from old data via transformation.
Plan Phase 4 (Code Switch): Update code to read from new schema, write to both. Gradually shift reads: 10% new, 90% old → 50/50 → 100% new.
Plan Phase 5 (Cleanup): Once all code reads new schema, remove old schema if safe. Keep removal as separate deployment after week of monitoring.