Data Architecture

Detailed instructions for Firestore schema design, data migration, and seeding.

When to use this skill

• •Use this when modifying server/src/seed.ts or src/types.
• •This is helpful for adding new collections, fields, or changing data relationships.
• •Use this when debugging data sync issues or "missing field" errors.

How to use it

1. Schema Integrity

• •Types First: Define the interface in src/types before writing any DB code.
• •Seed Truth: The seed.ts file is the ground truth. If a field exists in Types but not Seed, it's a bug.
• •No Deletions: Prefer deprecated: true flags over deleting fields to prevent runtime crashes on old clients.

2. Migrations & Seeding

• •Dry Run Mandate: Before updating seed.ts or running a migration, perform a comparison scan. Verify that incremental changes (e.g., adding a venue) do not accidentally wipe or regress existing records.
• •Idempotency: All migration scripts must be safe to run multiple times (check if data exists before writing).
• •Batching: Use db.batch() for multi-document updates to ensure atomicity.

3. Region

• •Firestore must be in us-west1.