Data Initialization and Cutover Plan (Greenfield Launch)

Context

This launch is greenfield from a prototype baseline with no production users.

Decision

• No legacy production data migration is required.
• Initialize production schema from db_schema_v1.sql.
• Seed only required system bootstrap data (SKUs, admin bootstrap account policy, baseline config).

Required Steps

1. Apply schema architecture/db_schema_v1.sql in dev, staging, prod.
2. Run bootstrap seed scripts for static reference data.
3. Validate schema invariants and health checks.
4. Enable APIs/workers only after contract and security gates pass.

If Prototype Data Is Needed for Demo/Testing

• Use one-time import into non-production environments only.
• Treat imported data as disposable.

Rollback Strategy

• For early greenfield stages: drop/recreate non-prod databases if needed.
• For production after launch: standard forward-only migration policy with tested rollback plans.

Notes

• This replaces legacy migration pressure and removes Phase 4 dependency on historical data transformation.
• For tenant/project ownership enforcement baseline, use reset-baseline execution in current dev stage:
• recreate schema with non-null ownership constraints,
• re-seed bootstrap data,
• do not implement temporary backfill code paths.