Reylo Migrate
Reylo Migrate is SQL-first schema modeling and migrations for Postgres, SQLite and MySQL.
- Schema in SQL: model your schema in
schema.sql - Your whole database: checks, indexes, triggers, functions, views and more, all tracked
- Preflight checks: each migration is tried on a shadow database before it runs
- Strict ordering: an optional lockfile,
reylo.lock, fixes the order migrations run in
Schema modeling
Your schema is plain SQL in one file, not a DSL like Prisma's.
create table users (
id integer primary key,
email text not null check (email like '%@%'),
verified_at timestamptz,
updated_at timestamptz not null default now()
);
create index users_unverified on users (id) where verified_at is null;
create function touch() returns trigger language plpgsql as $$
begin new.updated_at := now(); return new; end $$;
create trigger users_touch before update on users
for each row execute function touch();
create view verified_users as
select id, email from users where verified_at is not null;Change any of it, and reylo migrate new writes the migration. It tracks enums, domains, generated columns, materialized views, partitions, row level security policies, grants and comments too, on each database that has them.
Schemaless mode
Delete schema.sql, and your migrations are your schema.
A schema file reads best in review, but when an agent writes the DDL, migrations alone work too. With no schema to compare against, reylo migrate new drafts a migration from changes to your development database, or writes an empty one for you or your agent to fill:
$ reylo run 'alter table users add column bio text'
✓ alter table users add column bio text
$ reylo migrate new add_bio
✓ Wrote reylo/migrations/20261104140347_add_bio.sqlSee the database's schema at any time with reylo db introspect.
Migration ordering
An optional lockfile, reylo.lock, lists your migrations in order, each with a checksum chained to the one before:
schema: c7bf681b
migrations:
- 20261104120000_users.sql c86e9de4
- 20261104140347_add_bio.sql 78ad08c0Two branches that each add a migration conflict on it in git, before either reaches a database. For a team that writes many migrations, those conflicts add up.
Lock-free integrity
Delete reylo.lock, and migrations run in name order until reylo migrate lock writes it again.
Without a lock, a merged migration can sort before one a database already ran. Reylo replays both orders on a shadow database, and refuses where they end differently:
$ reylo migrate deploy
✗ Migration 20261105080000_no_nick.sql sorts before 20261105090000_nick.sql, which this database already ran, and the two orders end differently
+ a fresh database: column users.nick text
Give 20261105080000_no_nick.sql a later name, so that every database runs it after 20261105090000_nick.sql
Nothing was changed.A typical workflow
Write a migration, run it in development, then deploy it:
reylo migrate new add_bio # write the next migration, from schema.sql
reylo migrate dev # run it on your development database
reylo migrate deploy # run what the lock lists, when you deploy
reylo db push --watch # or, while you prototype, push schema.sql on every saveGet started
Reylo Migrate completes the Reylo toolkit. Reylo infers types from SQL queries, so you run them type-safe from TypeScript or Rust, with no codegen, ORM or query builder.
Start from your existing database:
npx reylo migrate initbrew install reylosql/tap/reylo
reylo migrate initcurl -fsSL https://reylo.dev/install.sh | sh
reylo migrate initIt writes your schema and a first migration, then the docs take it from there:
$ reylo migrate init
✓ Started reylo/ from postgres://localhost:5432/app
reylo/
├── schema.sql this database's schema
├── migrations/
│ └── init.sql the same, recorded as run here
└── reylo.lock the order the migrations run inReylo is open source, on GitHub.