Deployment Checklist

The Problem

Deployments fail because of forgotten steps — an env var not set in production, a database migration not run, a cache not cleared, monitoring not configured. Every team learns these lessons the hard way. Instead of relying on tribal knowledge and memory, generate a stack-specific checklist that covers every critical step.

The Prompt

Generate a comprehensive deployment checklist for my application. This will be used as a pre-deploy verification before pushing to production.

MY STACK:
- Frontend: [e.g., React, Astro, Next.js]
- Backend: [e.g., Node.js, Python/Django, Go]
- Database: [e.g., PostgreSQL, MongoDB, none]
- Hosting: [e.g., Vercel, AWS, Netlify, Docker/VPS]
- CI/CD: [e.g., GitHub Actions, GitLab CI, none]

DEPLOYMENT TYPE: [first deploy / routine update / major release / hotfix]

Generate a checklist organized by phase:

1. **Pre-Deploy Verification**
   - Tests, linting, type checking, build success

2. **Environment & Configuration**
   - Env vars, secrets, API keys, feature flags

3. **Database**
   - Migrations, backups, seed data, connection strings

4. **Security**
   - Dependencies audit, HTTPS, CORS, CSP headers, secrets rotation

5. **Performance**
   - Bundle size, caching headers, CDN, image optimization

6. **Monitoring & Rollback**
   - Error tracking, uptime monitoring, logging, rollback plan

7. **Post-Deploy Verification**
   - Smoke tests, critical user flows, monitoring dashboards

Mark each item as: [ ] Required or [~] Recommended

Example Output

# Deployment Checklist — Next.js + Vercel + PostgreSQL

## Pre-Deploy Verification
- [x] All tests pass (`npm test`)
- [x] TypeScript compiles without errors (`tsc --noEmit`)
- [x] Lint passes (`npm run lint`)
- [x] Build succeeds locally (`npm run build`)

## Environment & Configuration
- [ ] All env vars set in Vercel dashboard
- [ ] DATABASE_URL points to production database
- [ ] API keys are production keys (not dev/staging)
- [~] Feature flags reviewed for this release

When to Use

Generate a checklist before every production deployment, especially for first deploys to a new environment. For routine updates, use the generated checklist as a recurring template. Regenerate when your stack changes significantly.

Pro Tips

• Save and iterate — generate once, then refine based on actual deployment issues you encounter. Ask AI to “add a step for [specific thing that went wrong last time].”
• Make it team-specific — add “Our team uses Slack for deploy notifications and PagerDuty for alerts” to get integrated checklist items.
• Combine with CI — ask AI to convert the checklist into a GitHub Actions workflow that automates the verifiable steps.