Lindy Prod Checklist

Lindy Prod Checklist

Overview

Comprehensive production readiness checklist for Lindy AI deployments.

Prerequisites

• Completed development and testing
• Production Lindy account
• Deployment infrastructure ready

Production Checklist

Authentication & Security

[ ] Production API key generated
[ ] API key stored in secret manager (not env file)
[ ] Key rotation process documented
[ ] Different keys for each environment
[ ] Keys have appropriate scopes/permissions
[ ] Service accounts configured (not personal keys)

Agent Configuration

[ ] All agents tested with production-like data
[ ] Agent instructions reviewed and finalized
[ ] Tool permissions minimized (least privilege)
[ ] Timeout values appropriate for workloads
[ ] Error handling tested for all failure modes
[ ] Fallback behaviors defined

Monitoring & Observability

[ ] Logging configured and tested
[ ] Error alerting set up (PagerDuty/Slack/etc)
[ ] Usage metrics dashboards created
[ ] Rate limit alerts configured
[ ] Latency monitoring enabled
[ ] Cost tracking implemented

Performance & Reliability

[ ] Load testing completed
[ ] Rate limit handling implemented
[ ] Retry logic with exponential backoff
[ ] Circuit breaker pattern for failures
[ ] Graceful degradation defined
[ ] SLA targets documented

Compliance & Documentation

[ ] Data handling documented
[ ] Privacy review completed
[ ] Security review completed
[ ] Runbooks created for incidents
[ ] Escalation paths defined
[ ] On-call schedule set up

Implementation

Health Check Endpoint

// health/lindy.ts
import { Lindy } from '@lindy-ai/sdk';

export async function checkLindyHealth(): Promise<HealthStatus> {
  const lindy = new Lindy({ apiKey: process.env.LINDY_API_KEY });
  const start = Date.now();

  try {
    await lindy.users.me();
    const latency = Date.now() - start;

    return {
      status: latency < 1000 ? 'healthy' : 'degraded',
      latency,
      timestamp: new Date().toISOString(),
    };
  } catch (error: any) {
    return {
      status: 'unhealthy',
      error: error.message,
      timestamp: new Date().toISOString(),
    };
  }
}

Pre-Deployment Validation

async function preDeploymentCheck(): Promise<boolean> {
  const checks = {
    apiKey: !!process.env.LINDY_API_KEY,
    environment: process.env.LINDY_ENVIRONMENT === 'production',
    connectivity: false,
    agents: false,
  };

  const lindy = new Lindy({ apiKey: process.env.LINDY_API_KEY });

  try {
    await lindy.users.me();
    checks.connectivity = true;

    const agents = await lindy.agents.list();
    checks.agents = agents.length > 0;
  } catch (e) {
    // Failed checks
  }

  const passed = Object.values(checks).every(Boolean);
  console.log('Pre-deployment checks:', checks);
  console.log(`Status: ${passed ? 'PASSED' : 'FAILED'}`);

  return passed;
}

Output

• Complete production checklist
• Health check implementation
• Pre-deployment validation script
• Go/no-go criteria defined

Error Handling

Examples

Deployment Gate Script

#!/bin/bash
# deploy-gate.sh

echo "Running Lindy pre-deployment checks..."

npx ts-node scripts/pre-deployment-check.ts
if [ $? -ne 0 ]; then
  echo "Pre-deployment checks FAILED"
  exit 1
fi

echo "All checks passed. Proceeding with deployment."

Resources

• Lindy Production Guide
• SLA Information
• Support

Next Steps

Proceed to lindy-upgrade-migration for version upgrades.