supabase-engineer

name	supabase-engineer
description	Design and implement Supabase schema, migrations, Storage configuration, and typed client usage for the 3D media app. Use for new tables, views, RPC functions, storage buckets, and migration authoring.

Supabase Engineer

Use when

• Adding or modifying tables in libs/data/src/lib/migrations/*
• Writing or updating repository implementations in libs/data/src/lib/repositories/*
• Configuring Supabase Storage buckets or access policies
• Generating or updating database.types.ts after schema changes
• Debugging Supabase client errors or RLS denials

Schema overview

-- Core tables (see 001_initial_schema.sql)
products      (id, name, description, category, owner_id, created_at, updated_at)
conversions   (id, product_id, owner_id, source_asset_*, output_asset_*, status, error_message, ...)

-- Storage
bucket: media-assets  (public read, owner-scoped write)

Migration authoring rules

• Name migrations: NNN_description.sql (zero-padded, e.g. 002_add_tags.sql)
• Always include rollback comment block at the top
• Set DEFAULT NOW() on all timestamp columns
• Never use SERIAL — use gen_random_uuid() for PKs
• Add indexes for every FK and every column used in WHERE/ORDER BY
• Enable RLS on every new table immediately

Workflow

1. Read libs/data/src/lib/migrations/ to understand current schema.
2. Draft the migration SQL with RLS policies and indexes.
3. Update database.types.ts to reflect new table shape.
4. Update the relevant repository in libs/data/src/lib/repositories/.
5. Run migration: npx supabase db push (local) or apply via Supabase dashboard.

Supabase client

• Client is a singleton from libs/data/src/lib/supabase/client.ts
• Always use the typed Database generic: createClient<Database>(url, key)
• Never expose the service-role key client-side

Load only when needed

• Schema design guide
• Migration template
• RLS policy patterns
• Repository pattern guide