Database

stacks-database · Data layer · model-invoked

Connections, raw queries, SQL helpers and the four supported engines: SQLite, MySQL, PostgreSQL and DynamoDB. Covers what differs between them, which matters because local development is usually SQLite and production usually is not.

When to reach for it

• Configuring connections
• Running queries
• Migrations
• Seeding
• SQL helpers
• Using SQLite/MySQL/PostgreSQL/DynamoDB

Covers

@stacksjs/database, bun-query-builder, config/database.ts, the database/ migrations directory.

Inside the skill

The sections an agent reads once the skill loads.

• Key Paths
• Source Files
• Database Class (database.ts)
• Factory Functions (database.ts)
• Driver Configuration (driver-config.ts)
• Global db Instance (utils.ts)
• SQL Template Tag (types.ts)
• SQL Dialect Helpers (sql-helpers.ts)
• Connection Defaults (defaults.ts)
• DatabaseOptions Type (database.ts)
• Connection Types (driver-config.ts)
• Migrations (migrations.ts)
• Seeding (seeder.ts)
• Validator Type Guards (validators.ts)
• DynamoDB Support (drivers/dynamodb.ts)
• Re-exports from bun-query-builder
• Compatibility Type Aliases (types.ts)
• CLI Commands
• config/database.ts Shape
• config/query-builder.ts (Query Builder Config)
• Gotchas

Where the code lives

• Database package: storage/framework/core/database/src/
• Configuration: config/database.ts
• QB config: config/query-builder.ts
• Migrations: database/migrations/ (96+ migration files, .sql format)
• QB state: .qb/
• ORM: storage/framework/orm/

Using it

This one is model-invoked. Your agent reaches for it on its own when the task matches, and you can also call it by name:

/stacks-database

Source: stacks-database/SKILL.md. Shadow it for one project with app/Skills/stacks-database/SKILL.md, then re-run buddy setup:ai. See Writing your own.