Cloudflare Workers

Overview

Cloudflare Workers enables building and deploying applications at the edge with sub-millisecond cold starts. The platform leverages the Workers runtime alongside storage services like KV, D1, R2, Durable Objects, and Queues to build globally distributed, low-latency applications.

Instructions

• When asked to create a Worker, scaffold with wrangler init using ES Module syntax (export default { fetch }) and set compatibility_date in wrangler.toml.
• When configuring storage, recommend KV for read-heavy key-value caching, D1 for relational data with SQL, R2 for S3-compatible object storage with zero egress fees, and Durable Objects for strongly consistent state coordination.
• When setting up local development, use wrangler dev with hot reload and local KV/D1/R2 simulation.
• When deploying, use wrangler deploy and configure routes, bindings, and build settings in wrangler.toml.
• When managing secrets, use wrangler secret put KEY_NAME and type bindings with an Env interface.
• When optimizing performance, leverage the Cache API (caches.default), Smart Placement, streaming responses with TransformStream, and HTMLRewriter for HTML transformation.
• When handling background work, use ctx.waitUntil() for fire-and-forget async tasks like analytics or logging.
• When building AI features, use Workers AI for edge inference, AI Gateway for multi-provider management, and Vectorize for RAG pipelines.

Examples