prisma-cli

Run prisma migrate dev --name your_migration_name to create and apply a migration from schema changes. It uses a shadow database for drift detection. Afterwards, run prisma generate and prisma db seed explicitly if you need refreshed client output or seed data.

db push syncs the schema directly without creating migration files or tracking history, making it faster for prototyping and required for MongoDB. migrate dev creates versioned migration files with rollback capability and drift detection, suited for team development.

Use prisma migrate deploy, which applies pending migrations without prompts, shadow databases, or drift detection, making it safe for CI/CD pipelines. Run prisma migrate status first to verify pending migrations before deploying.

Prisma detects AI agent environments and blocks destructive commands like migrate reset and db push --force-reset until explicit user consent is given. The agent must explain the data-loss impact, obtain consent, and pass the exact consent text via PRISMA_USER_CONSENT_FOR_DANGEROUS_AI_ACTION.

Yes, but you must run it with bunx --bun so Prisma uses the Bun runtime instead of falling back to Node.js due to the CLI shebang. This applies to commands like bunx --bun prisma init and bunx --bun prisma generate.

Run prisma db pull after configuring the datasource URL in prisma.config.ts. It reads tables, columns, and relations into schema.prisma. Use --print to preview first, since pulling overwrites the existing schema file.