db-migration

Database Migration: $ARGUMENTS

Skill bundle root

Every bundled helper below runs through the skill bundle root, never a bare scripts/… path (the working directory is the user's project, where scripts/ is absent or belongs to someone else):

SKILL_DIR="${MB_SKILLS_ROOT:-${SKILL_DIR:-$HOME/.claude/skills/memory-bank}}"
[ -f "$SKILL_DIR/scripts/_lib.sh" ] || { echo "mb: skill bundle not found at $SKILL_DIR — set MB_SKILLS_ROOT" >&2; exit 2; }

0. Validate arguments

If $ARGUMENTS is empty, stop and ask the user what the migration should change.

1. Stack detection

eval "$(bash "$SKILL_DIR"/scripts/mb-metrics.sh)"
# Uses: stack to narrow the tooling search

2. Detect the migration tool

# golang-migrate / goose / Atlas (Go)
ls migrations/ db/migrations/ 2>/dev/null
grep -r "golang-migrate\|pressly/goose\|ariga/atlas" go.mod 2>/dev/null

# Alembic (Python)
ls alembic/ 2>/dev/null; cat alembic.ini 2>/dev/null

# Prisma (Node.js)
ls prisma/schema.prisma 2>/dev/null

# Sequelize / Knex (Node.js)
ls sequelize/ migrations/ knexfile.js knexfile.ts 2>/dev/null
grep -l "sequelize\|knex" package.json 2>/dev/null

# Diesel (Rust)
ls diesel.toml 2>/dev/null
grep -l "diesel" Cargo.toml 2>/dev/null

# SQLx (Rust)
ls migrations/ 2>/dev/null
grep -l "sqlx" Cargo.toml 2>/dev/null

# Flyway (Java/JVM)
ls flyway.conf db/migration/ src/main/resources/db/migration/ 2>/dev/null

# Liquibase (Java/JVM)
ls liquibase.properties changelog*.xml db/changelog/ 2>/dev/null

# Entity Framework Core (.NET)
ls Migrations/ 2>/dev/null
grep -l "EntityFrameworkCore" *.csproj 2>/dev/null

# Plain SQL files
find . -name "*.sql" -path "*/migrat*" 2>/dev/null | head -20

If none of the above matches, ask the user which migration tool the project uses before writing anything.

3. Universal requirements

Every migration must follow these rules, regardless of tool:

• Both up and down (or equivalent forward / rollback). down must fully reverse up.
• Destructive operations (DROP TABLE, DROP COLUMN, TRUNCATE, DELETE FROM without WHERE) require explicit y/N confirmation before the file is written — not just before it is applied. Default = No.
• Schema migrations and data migrations must be in separate files.
• Index creation uses CONCURRENTLY on PostgreSQL when the tool supports it (golang-migrate -disable-create-hash-comments, Alembic with op.create_index(..., postgresql_concurrently=True), etc.).
• Idempotent wrappers (IF NOT EXISTS, IF EXISTS) on DDL where the dialect supports them.
• Backward compatibility: the previous application version must still function against the new schema for the duration of a rolling deploy.