cloudflare-workers_skill

Overview

This skill helps you develop, test, and deploy Cloudflare Workers for serverless edge applications. It focuses on rapid iteration with Wrangler, bindings (KV, D1, R2, Durable Objects), cron/queue handlers, and framework adapters to run code globally with low latency. Use it to build APIs, full-stack apps, edge middleware, background jobs, and real-time systems.

How this skill works

The skill inspects project files and trigger phrases (wrangler.toml, worker.ts, worker.js, wrangler, cloudflare workers) to guide setup, local development, and deployment. It explains runtime limits, handler types (fetch, scheduled, queue), and binding configuration so you can wire KV, D1, R2, and secrets into your Worker. It also provides patterns for testing, CORS, authentication, and framework integration for Next.js, Remix, Astro, SvelteKit, and Hono.

When to use it

• Build low-latency HTTP APIs and WebSocket services at the edge
• Serve full-stack applications with static assets and server-side rendering
• Run edge middleware: auth, rate limiting, A/B testing, and routing
• Schedule background jobs or consume queues with cron and queue handlers
• Integrate serverless functions with D1, KV, R2, Durable Objects, or Workers AI
• Deploy container workloads or durable workflows requiring global execution

Best practices

• Use Wrangler CLI and environments (staging/production) to manage deployments and secrets
• Prefer bindings (KV, D1, R2) rather than embedding credentials; store secrets via wrangler secret put
• Keep handlers idempotent and use ctx.waitUntil for post-response work to reduce latency
• Respect runtime limits: optimize for CPU/memory and break long tasks into durable workflows or containers
• Add comprehensive tests (Vitest with cloudflare:test) and CI/CD deployment pipelines
• Use top-level env import for module-level initialization where supported to avoid prop-drilling

Example use cases

• REST or GraphQL API serving global users with <100ms latency per region
• Edge middleware performing authentication and A/B routing before hitting origin
• Background workers that process webhooks, run ETL jobs, or consume queues on schedule
• Full-stack apps serving static assets from R2 with SSR via framework adapters
• Real-time services using WebSockets or Workers AI for RAG chatbots