Skip to main content

​
Pre-Launch Checklist

Before launching your AI agent to production, ensure you’ve completed these essential steps.
  • Generate production API key from Nevermined App
  • Store API key securely in environment variables
  • Switch from testing to production environment
  • Verify connection to production services
const payments = Payments.getInstance({
  nvmApiKey: process.env.NVM_API_KEY,
  environment: 'production'
})
  • Complete agent metadata (name, description, tags, image)
  • Define all API endpoints correctly
  • Specify open endpoints that don’t require payment
  • Test agent registration on production
const agentMetadata = {
  name: 'Production AI Agent',
  tags: ['production', 'ai', 'v1.0'],
  dateCreated: new Date(),
  description: 'Production-ready AI service',
  image: 'https://yourdomain.com/agent-logo.png'
}
  • Set appropriate pricing for your service
  • Choose between ERC-20 tokens or fiat (Stripe)
  • Configure credit allocation (fixed, dynamic, or time-based)
  • Test payment flow end-to-end
  • Set up multiple tiers if needed
// Production pricing
const priceConfig = getERC20PriceConfig(
  25_000_000n, // 25 USDC
  '0xA0b86a33E6441c41F4F2B8Bf4F2B0f1B0F1C1C1C',
  process.env.BUILDER_WALLET_ADDRESS
)
  • Implement request validation on all endpoints
  • Add rate limiting to prevent abuse
  • Set up HTTPS/TLS for all endpoints
  • Implement proper error handling
  • Add request logging for auditing
// Validation middleware
app.use(neverminedAuth(planId, agentId))
app.use(rateLimiter)
app.use(helmet())
  • Implement caching for agent/plan data
  • Set up retry logic for transient failures
  • Optimize database queries
  • Configure appropriate timeouts
  • Load test your endpoints
// Cache configuration
const cache = new NodeCache({ stdTTL: 300 }) // 5 min TTL
  • Set up application monitoring (APM)
  • Configure error tracking (Sentry, etc.)
  • Implement revenue tracking
  • Set up health check endpoints
  • Configure alerting for critical issues
app.get('/health', async (req, res) => {
  const health = await checkSystemHealth()
  res.status(health.isHealthy ? 200 : 503).json(health)
})
  • Document API endpoints and parameters
  • Create integration guide for subscribers
  • Provide code examples in multiple languages
  • Document error codes and responses
  • Create FAQ section
  • Unit tests for all critical functions
  • Integration tests for payment flows
  • End-to-end tests for user journeys
  • Load testing for expected traffic
  • Security testing for vulnerabilities
npm test
npm run test:integration
npm run test:load

​
Production Configuration

​
Environment Variables

Ensure all required environment variables are set: 
# Required
NVM_API_KEY=your_production_api_key
BUILDER_ADDRESS=your_wallet_address
NODE_ENV=production

# Recommended
REDIS_URL=redis://your-redis-instance
SENTRY_DSN=your_sentry_dsn
LOG_LEVEL=info
RATE_LIMIT_MAX=100
RATE_LIMIT_WINDOW_MS=900000

​
Infrastructure Requirements

Server Requirements

  • HTTPS/TLS enabled
  • Auto-scaling configured
  • Load balancer set up
  • CDN for static assets

Database & Storage

  • Database backups configured
  • Redis for caching
  • Log aggregation service
  • Metrics storage

​
Launch Day Checklist

Complete these final checks on launch day to ensure a smooth deployment.

​
Final Verification

  • All tests passing in CI/CD pipeline
  • Production API key working correctly
  • Payment flow tested with real transactions
  • Monitoring dashboards configured
  • Support channels ready

​
Deployment Steps

  1. Deploy to Production 
    npm run build
npm run deploy:production
  2. Verify Deployment 
    curl https://your-api.com/health
  3. Test Live Endpoints 
    const agent = await payments.getAgent(productionAgentId)
console.log('Agent live:', agent.metadata.name)
  4. Monitor Initial Traffic
    • Check error rates
    • Monitor response times
    • Verify payment processing
    • Watch for rate limit hits

​
Post-Launch Tasks

​
First 24 Hours

  • Monitor error rates and fix critical issues
  • Check payment processing success rate
  • Review performance metrics
  • Respond to user feedback
  • Update documentation based on user questions

​
First Week

  • Analyze usage patterns
  • Optimize based on real-world data
  • Plan feature updates
  • Review security logs
  • Calculate initial revenue metrics

​
Troubleshooting Common Issues

Problem: Getting “Invalid API key” errors in productionSolution:
  • Verify you’re using production API key
  • Check environment variable is set correctly
  • Ensure API key hasn’t expired
  • Generate new key if needed
Problem: Payments failing or not processingSolution:
  • Verify wallet address is correct
  • Check token contract addresses
  • Ensure sufficient gas for transactions
  • Test with smaller amounts first
Problem: Slow response times or timeoutsSolution:
  • Implement caching for repeated queries
  • Check database query performance
  • Scale infrastructure if needed
  • Review and optimize code paths

​
Support Resources

If you encounter issues during launch:
Remember to celebrate your launch! You’re contributing to the AI-native economy by enabling sustainable monetization for AI services.