cloudflare-deploy

Authenticate with wrangler login or set the CLOUDFLARE_API_TOKEN environment variable, then run wrangler deploy. Verify authentication first with npx wrangler whoami before deploying.

Use KV for key-value config and caching, D1 for relational SQLite data, R2 for S3-compatible object storage, Queues for async messaging, and Vectorize for embeddings. The skill's decision tree maps each storage need to the right product.

Use Workers for serverless functions and edge logic, and Pages for full-stack web apps with Git-based deploys. The skill routes compute needs through a decision tree covering both options plus Durable Objects and Containers.

Yes, the skill includes references for both Terraform and Pulumi providers covering zones, DNS, Workers, KV, R2, D1, and Pages. Terraform v5 renamed many resources, so check the migration notes in the terraform reference.

Sandboxed environments may block outbound network calls needed for deployment. Rerun the deploy command with escalated permissions using sandbox_permissions=require_escalated to allow the network access.

Do not use it for deploying to Vercel, Netlify, or Render, since those platforms have their own dedicated skills. It is scoped exclusively to the Cloudflare platform and its services.