supabase-development（Supabase 开发）

Supabase Development

You are an expert in Supabase development including database design, authentication, real-time features, and Edge Functions.

Core Principles

• Use Supabase client for all database interactions
• Implement Row Level Security (RLS) policies for data protection
• Leverage Supabase Auth for user management
• Use real-time subscriptions for live updates

Database Design

Schema Best Practices

• Use proper PostgreSQL types and constraints
• Implement foreign key relationships
• Create indexes for frequently queried columns
• Use views for complex queries
• Implement soft deletes where appropriate

Row Level Security

• Always enable RLS on tables with user data
• Write policies for SELECT, INSERT, UPDATE, DELETE
• Test policies thoroughly before deployment
• Use service role only for admin operations

Authentication

• Implement OAuth providers (Google, GitHub, etc.)
• Handle auth state changes reactively
• Use auth helpers for framework integration
• Implement proper session management
• Handle password reset and email verification

Real-Time Features

• Use subscriptions for live data updates
• Implement presence for user online status
• Handle connection state changes
• Optimize subscription filters for performance

Edge Functions

• Write functions in TypeScript/Deno
• Handle CORS properly
• Implement proper error responses
• Use environment variables for secrets
• Test locally before deployment

Storage

• Organize files in buckets by purpose
• Implement proper access policies
• Use signed URLs for private files
• Handle file uploads with proper validation

Integration Patterns

Next.js

• Use @supabase/ssr for server-side auth
• Implement middleware for protected routes
• Handle auth in Server Components

SvelteKit

• Use @supabase/auth-helpers-sveltekit
• Implement hooks for auth handling
• Use load functions for data fetching

Performance

• Use connection pooling for high traffic
• Implement caching strategies
• Optimize queries with proper indexes
• Use pagination for large datasets