Deployment manager

Deployment Manager

Context

Use this skill to deploy projects, configure CI/CD, manage environment variables, monitor deployment health, or execute rollbacks. Covers GitHub Pages, Vercel, and Netlify.

Trigger phrases: "deploy to Vercel/Netlify/GitHub Pages," "configure CI/CD," "rollback deployment," "monitor uptime," "environment variables."

---

Instructions

Step 1: Pre-Deployment Checklist

1. Build locally: npm run build — no errors, reasonable bundle size.
2. Test: npm run test -- --coverage + npm run lint — all pass.
3. Environment: create .env.example (document vars, no secrets), verify .gitignore includes .env*.
4. Git: all changes committed, correct branch (main), no untracked files.

Step 2: Platform Deployment

GitHub Pages

5. Static sites only. Add .github/workflows/deploy.yml with actions/deploy-pages.
6. Set output directory (dist/build). Settings → Pages → Source: GitHub Actions.
7. For SPAs: add 404.html redirecting to index.html. Set base in build config.

Vercel

8. Connect repo or npx vercel. Auto-detects framework; override build/output commands if needed.
9. Set env vars in Project Settings → Environment Variables (Production/Preview/Dev).
10. Custom domain: add in Settings → Domains. DNS: CNAME to cname.vercel-dns.com or A record 76.76.21.21. SSL auto-provisioned.

Netlify

11. npm i -g netlify-cli && netlify init. Set build command and publish directory.
12. Redirects in netlify.toml: [[redirects]] from="/*" to="/index.html" status=200.
13. Set env vars in Site Settings → Environment.

Step 3: SSL & CDN

14. All platforms auto-provision SSL (Let's Encrypt). Verify HTTPS is active.
15. Cache headers for hashed assets: Cache-Control: public, max-age=31536000, immutable.
16. Lighthouse audit: target Performance > 90, Accessibility > 90. Core Web Vitals: LCP < 2.5s, FID < 100ms, CLS < 0.1.

Step 4: Monitoring

17. Uptime: UptimeRobot (free, 1-5 min checks) or BetterUptime.
18. Errors: Sentry (free tier). Add Sentry.init({ dsn }) in app entry.
19. Alerts: email on > 5 errors/hour, Slack on > 10 errors/hour.
20. Performance: Vercel Speed Insights or Netlify Analytics. Track load times, API latency, bundle size trends.

Step 5: Update & Rollback

21. Update flow: Pull → Install → Test → Build → Push to main → CI runs → Verify → Smoke test → Monitor 15 min.
22. Rollback by platform:
    • Vercel: Dashboard → Deployments → "..." → Promote to Production (30s)
    • Netlify: Dashboard → Deploys → Rollback previous
    • GitHub Pages: git revert [hash] → push → Actions redeploys
23. Smoke test: homepage loads, nav works, key flows function, no console errors, assets load, APIs respond.

---

Constraints

• NEVER deploy locally to production for team projects. Always use CI/CD.
• NEVER commit secrets. Use platform UI or encrypted secrets.
• ALWAYS test on preview before promoting to production.
• ALWAYS have a rollback plan before deploying.
• NEVER skip the post-deploy smoke test.
• Environment variables MUST be documented in .env.example.
• SSL MUST be active on production.
• Deploy logs MUST be checked for warnings even on successful builds.

---

Examples

Example 1: React SPA → Vercel

Project: React + TypeScript + Vite portfolio
1. npm run build → dist/
2. Push to GitHub main → Vercel auto-deploys
3. Env vars: VITE_API_URL, VITE_CONTACT_EMAIL
4. Domain: portfolio.jane.dev → CNAME cname.vercel-dns.com
5. SSL auto, Lighthouse 96/98/100
6. Sentry for errors, UptimeRobot 5-min pings
Rollback: Vercel dashboard → Promote previous deploy

Example 2: Docs Site → GitHub Pages

Project: Astro + Tailwind documentation
1. .github/workflows/deploy.yml → actions/deploy-pages
2. Push main → Actions builds and deploys
3. Domain: docs.project.com → CNAME github.io
4. Enable Enforce HTTPS toggle
5. 404.html for SPA routing fallback
6. UptimeRobot 1-min checks → Slack alerts
7. Weekly Lighthouse via cron GitHub Action
Rollback: git revert [hash] → push → Actions redeploys