Overview

Approach 2: Multi-Repl (Staging + Production)

For teams that need a staging environment:

Repl 1: my-app-staging → Autoscale deployment → staging.replit.app
Repl 2: my-app-prod   → Reserved VM deployment → app.example.com

Each Repl has its own:
- Secrets (different API keys per environment)
- PostgreSQL database (separate data)
- Deployment configuration
- GitHub branch (staging → staging, main → production)

Instructions

Step 1: Environment Detection

// src/config/environment.ts
type Environment = 'development' | 'staging' | 'production';

export function detectEnvironment(): Environment {
  // Replit deployment context
  if (process.env.REPL_DEPLOYMENT) {
    // Check if this is the staging Repl
    if (process.env.REPL_SLUG?.includes('staging')) return 'staging';
    return 'production';
  }

  // Workspace "Run" context
  if (process.env.REPL_SLUG) return 'development';

  // Fallback to NODE_ENV
  const env = process.env.NODE_ENV || 'development';
  if (env === 'production') return 'production';
  if (env === 'staging') return 'staging';
  return 'development';
}

export const ENV = detectEnvironment();
export const IS_PROD = ENV === 'production';

Step 2: Environment-Specific Configuration

// src/config/index.ts
import { ENV, IS_PROD } from './environment';

const configs = {
  development: {
    logLevel: 'debug',
    corsOrigins: ['*'],
    rateLimit: { windowMs: 60000, max: 1000 },
    cache: { ttlMs: 5000 },
    features: { debugEndpoints: true },
  },
  staging: {
    logLevel: 'info',
    corsOrigins: ['https://staging.example.com'],
    rateLimit: { windowMs: 60000, max: 200 },
    cache: { ttlMs: 30000 },
    features: { debugEndpoints: true },
  },
  production: {
    logLevel: 'warn',
    corsOrigins: ['https://app.example.com'],
    rateLimit: { windowMs: 60000, max: 100 },
    cache: { ttlMs: 300000 },
    features: { debugEndpoints: false },
  },
};

export const config = {
  env: ENV,
  port: parseInt(process.env.PORT || '3000'),
  ...configs[ENV],
  db: {
    // DATABASE_URL auto-switches between dev and prod on Replit
    url: process.env.DATABASE_URL,
  },
};

// Validate production secrets
if (IS_PROD) {
  const required = ['DATABASE_URL', 'JWT_SECRET'];
  const missing = required.filter(k => !process.env[k]);
  if (missing.length) {
    console.error(`FATAL: Missing production secrets: ${missing.join(', ')}`);
    process.exit(1);
  }
}

Step 3: Separate Secrets Per Environment

For Single Repl (dev/prod separation):
- All secrets set once in Secrets tab
- DATABASE_URL auto-switches (Replit manages)
- Same JWT_SECRET for both (or use REPL_IDENTITY for dev)

For Multi-Repl (staging + prod):
- Each Repl has its own Secrets tab
- staging Repl: JWT_SECRET=staging-secret, API_KEY=test-key
- production Repl: JWT_SECRET=prod-secret, API_KEY=live-key

Account-level secrets (shared across all Repls):
- Settings > Secrets > Account secrets
- Useful for: monitoring tokens, shared infrastructure keys

Step 4: GitHub Branch Strategy (Multi-Repl)

Repository setup:
- main branch → connected to production Repl
- staging branch → connected to staging Repl
- feature branches → PR to staging first

Workflow:
1. Feature branch → PR to staging
2. Tests pass + review → merge to staging
3. Staging Repl auto-deploys → verify staging
4. Staging → PR to main
5. Merge → production Repl auto-deploys

# .github/workflows/deploy.yml