Imported from jeremylongshore/tons-of-skills-marketplace (plugins/saas-packs/klaviyo-pack/skills/klaviyo-deploy-integration/SKILL.md). Install upstream with npx skills add jeremylongshore/tons-of-skills-marketplace --skill klaviyo-deploy-integration. Copyright stays with the author (MIT).

Klaviyo Deploy Integration

Overview

Deploy Klaviyo-powered applications to Vercel, Fly.io, and Google Cloud Run with proper secrets management and health checks. Every platform follows the same shape — store the private key + webhook secret, wire a config file, deploy, verify. The lean skeleton lives here; full per-platform recipes live in references/platform-deployments.md.

Prerequisites

• Klaviyo production API key (pk_*)
• Platform CLI installed (vercel, fly, or gcloud)
• Application tested with klaviyo-api SDK
• klaviyo-prod-checklist completed

Instructions

The workflow is identical across platforms; only the CLI verbs change. Read the target platform's section in references/platform-deployments.md, then:

1. Store secrets — inject KLAVIYO_PRIVATE_KEY and KLAVIYO_WEBHOOK_SIGNING_SECRET via the platform's secret store (vercel env add, fly secrets set, or gcloud secrets create). Never commit these to the repo.
2. Write the platform config — use Write/Edit to create vercel.json, fly.toml, or a Dockerfile that binds the secrets and exposes a health path. See the reference for the exact file contents.
3. Add the universal health check — expose src/health.ts (identical on all platforms) at the path each platform probes (/api/health on Vercel, /health on Fly.io and Cloud Run).
4. Deploy — run vercel --prod, fly deploy, or gcloud run deploy.
5. Verify — curl the health endpoint and confirm services.klaviyo.connected is true.

Vercel skeleton (first example)

vercel env add KLAVIYO_PRIVATE_KEY production          # paste pk_*** when prompted
vercel env add KLAVIYO_WEBHOOK_SIGNING_SECRET production
# configure vercel.json (see reference), then:
vercel --prod
curl -s https://your-app.vercel.app/api/health | jq '.services.klaviyo'

Fly.io (fly secrets set + fly.toml + fly deploy) and Cloud Run (gcloud secrets create + Dockerfile + gcloud run deploy --set-secrets) follow the same five steps — full commands and config files are in the reference walkthroughs.

Output

• Application deployed with Klaviyo secrets configured
• Health check endpoint verifying Klaviyo connectivity
• Webhook endpoint with HMAC signature verification
• Platform-specific best practices applied

Error Handling

Examples

Deploy to Fly.io with a health check. Set the two secrets, deploy, and confirm connectivity:

fly secrets set KLAVIYO_PRIVATE_KEY=pk_*** \
  KLAVIYO_WEBHOOK_SIGNING_SECRET=whsec_***
fly deploy
curl -s https://my-klaviyo-app.fly.dev/health | jq '.services.klaviyo'
# → { "connected": true, "latencyMs": 142 }

Verify a webhook signature (HMAC-SHA256). Every platform's webhook route must timing-safe-compare the klaviyo-webhook-signature header against an HMAC of the raw body keyed with the signing secret — a 401 otherwise. The full Vercel handler, vercel.json, fly.toml, the Cloud Run Dockerfile, and the universal src/health.ts are in references/platform-deployments.md.

For webhook event handling beyond signature verification, see the klaviyo-webhooks-events skill.

Resources

• Vercel Environment Variables
• Fly.io Secrets
• Cloud Run Secrets
• Klaviyo API Reference
• Full platform walkthroughs — verbatim commands, config files, and code for all three platforms