This skill provides comprehensive guidance for developing Cloudflare Workers, including runtime APIs, fetch event handlers, request/response handling, bindings usage, and common development patterns. Use this skill when implementing Workers code, designing Workers architecture, or working with the Workers runtime environment.
Cloudflare Workers run on the V8 JavaScript engine at the edge. Workers use the Service Worker API with extensions specific to the Cloudflare platform.
- Isolate-based architecture: Each request runs in an isolated V8 context, not a container
- Fast cold starts: Sub-millisecond startup time due to isolate architecture
- Automatic scaling: No configuration needed, scales to millions of requests
- Global deployment: Code runs at 300+ Cloudflare data centers worldwide
- Standards-based: Uses Web APIs (fetch, Request, Response, Headers, etc.)
Workers execute on incoming requests:
- Request arrives at Cloudflare edge
- Worker isolate spawns (or reuses existing)
- Fetch event handler executes
- Response returns to client
- Isolate may persist for subsequent requests
The fetch event handler is the entry point for Workers:
export default {
async fetch(request, env, ctx) {
// request: Request object
// env: Bindings and environment variables
// ctx: Execution context with waitUntil() and passThroughOnException()
return new Response('Hello World');
}
};
request: Incoming
Request
object (Web API standard)
request.url
- Full URL stringrequest.method
- HTTP method (GET, POST, etc.)request.headers
- Headers objectrequest.body
- ReadableStream of request bodyrequest.cf
- Cloudflare-specific properties
env: Environment object containing bindings
- KV namespaces:
env.MY_KV
- D1 databases:
env.MY_DB
- R2 buckets:
env.MY_BUCKET
- Durable Objects:
env.MY_DO
- Secrets:
env.MY_SECRET
- Environment variables:
env.MY_VAR
ctx: Execution context
ctx.waitUntil(promise)
- Extend execution lifetime for async tasksctx.passThroughOnException()
- Pass request to origin if Worker throws
Must return a
Response
object or a Promise that resolves to a Response:
// Direct return
return new Response('Hello', { status: 200 });
// Async return
return await fetch('https://api.example.com');
// With headers
return new Response(JSON.stringify({ ok: true }), {
status: 200,
headers: {
'Content-Type': 'application/json'
}
});
Common routing patterns for multi-route Workers:
export default {
async fetch(request, env, ctx) {
const url = new URL(request.url);
// Path-based routing
if (url.pathname === '/api/users') {
return handleUsers(request, env);
}
if (url.pathname.startsWith('/api/')) {
return handleAPI(request, env);
}
// Method-based routing
if (request.method === 'POST') {
return handlePost(request, env);
}
// Default
return new Response('Not Found', { status: 404 });
}
};
See
examples/fetch-handler-patterns.js
for complete routing examples.
Access request properties:
const url = new URL(request.url);
const method = request.method;
const headers = request.headers.get('Authorization');
const cookies = request.headers.get('Cookie');
// Cloudflare-specific properties
const country = request.cf?.country;
const colo = request.cf?.colo; // Data center code
Parse request bodies based on content type:
// JSON
const data = await request.json();
// Form data
const formData = await request.formData();
const field = formData.get('fieldName');
// Text
const text = await request.text();
// Array buffer
const buffer = await request.arrayBuffer();
// Text response
return new Response('Hello World');
// JSON response
return new Response(JSON.stringify({ message: 'Success' }), {
headers: { 'Content-Type': 'application/json' }
});
// HTML response
return new Response('<h1>Hello</h1>', {
headers: { 'Content-Type': 'text/html' }
});
// Status codes
return new Response('Not Found', { status: 404 });
return new Response('Created', { status: 201 });
// Set headers
const headers = new Headers({
'Content-Type': 'application/json',
'Cache-Control': 'max-age=3600',
'X-Custom-Header': 'value'
});
return new Response(body, { headers });
// CORS headers
const corsHeaders = {
'Access-Control-Allow-Origin': '*',
'Access-Control-Allow-Methods': 'GET, POST, OPTIONS',
'Access-Control-Allow-Headers': 'Content-Type'
};
// 301 permanent redirect
return Response.redirect('https://example.com', 301);
// 302 temporary redirect
return Response.redirect('https://example.com', 302);
Bindings provide access to Cloudflare resources and are configured in
wrangler.toml
or
wrangler.jsonc
.
- KV: Key-value storage for static content
- D1: SQLite database for relational data
- R2: Object storage for large files
- Durable Objects: Stateful coordination and real-time features
- Queues: Message queuing for async processing
- Vectorize: Vector database for embeddings
- Workers AI: AI inference and embeddings
- Service bindings: Call other Workers
- Environment variables: Configuration values
- Secrets: Sensitive credentials
See
references/bindings-guide.md
for complete binding configuration and usage patterns.
KV (Key-Value):
// Read
const value = await env.MY_KV.get('key');
const json = await env.MY_KV.get('key', 'json');
// Write
await env.MY_KV.put('key', 'value');
await env.MY_KV.put('key', JSON.stringify(data), {
expirationTtl: 3600 // Expire in 1 hour
});
D1 (Database):
// Query
const result = await env.MY_DB.prepare(
'SELECT * FROM users WHERE id = ?'
).bind(userId).first();
// Insert
await env.MY_DB.prepare(
'INSERT INTO users (name, email) VALUES (?, ?)'
).bind(name, email).run();
R2 (Object Storage):
// Read
const object = await env.MY_BUCKET.get('file.txt');
const text = await object.text();
// Write
await env.MY_BUCKET.put('file.txt', 'content');
export default {
async fetch(request, env, ctx) {
try {
// Your logic here
const result = await processRequest(request, env);
return new Response(JSON.stringify(result), {
headers: { 'Content-Type': 'application/json' }
});
} catch (error) {
console.error('Error:', error);
return new Response(JSON.stringify({
error: error.message
}), {
status: 500,
headers: { 'Content-Type': 'application/json' }
});
}
}
};
function errorResponse(message, status = 500) {
return new Response(JSON.stringify({ error: message }), {
status,
headers: { 'Content-Type': 'application/json' }
});
}
// Usage
if (!apiKey) {
return errorResponse('API key required', 401);
}
See
examples/error-handling.js
for comprehensive error handling patterns.
Use
ctx.waitUntil()
to perform background tasks that extend beyond the response:
export default {
async fetch(request, env, ctx) {
// Respond immediately
const response = new Response('Request received');
// Continue processing in background
ctx.waitUntil(
logRequest(request, env)
);
return response;
}
};
async function logRequest(request, env) {
await env.MY_DB.prepare(
'INSERT INTO logs (url, timestamp) VALUES (?, ?)'
).bind(request.url, Date.now()).run();
}
Important:
waitUntil
extends execution but doesn't guarantee completion. Use for non-critical tasks like logging, analytics, or cache warming.
- Minimize CPU time: Workers are billed by CPU time, keep processing lean
- Use edge caching: Cache responses at the edge when possible
- Parallel requests: Use
Promise.all()
for concurrent operations
- Avoid blocking: Don't use synchronous APIs or long computations
- Validate inputs: Always validate and sanitize user input
- Use secrets: Store sensitive data in secrets, not environment variables
- CORS properly: Configure CORS headers correctly for browser requests
- Rate limiting: Implement rate limiting for public APIs
- Console logging: Use
console.log()
for debugging (visible in wrangler tail
)
- Local testing: Test with
wrangler dev
before deploying
- Real-time logs: Use
wrangler tail
to see production logs
- Error handling: Catch and log errors with context
- Separate concerns: Split routing, business logic, and data access
- Reusable functions: Create helper functions for common operations
- Type safety: Use TypeScript for better IDE support and fewer bugs
- Environment-aware: Use bindings through
env
, not globals
Workers support standard Web APIs and Cloudflare-specific extensions.
- fetch() - HTTP requests (Web API standard)
- Request/Response - HTTP primitives
- Headers - Header manipulation
- URL - URL parsing and construction
- URLSearchParams - Query string handling
- ReadableStream/WritableStream - Streaming data
- crypto - Cryptographic operations
- TextEncoder/TextDecoder - Text encoding
- atob/btoa - Base64 encoding
- request.cf - Cloudflare request properties (country, colo, etc.)
- HTMLRewriter - HTML parsing and transformation
- Cache API - Edge caching control
- scheduled() - Cron Triggers (in addition to fetch)
See
references/runtime-apis.md
for complete API documentation and examples.
Workers fully support TypeScript with official type definitions:
export interface Env {
MY_KV: KVNamespace;
MY_DB: D1Database;
MY_BUCKET: R2Bucket;
MY_SECRET: string;
}
export default {
async fetch(
request: Request,
env: Env,
ctx: ExecutionContext
): Promise<Response> {
// Type-safe access to bindings
const value = await env.MY_KV.get('key');
return new Response(value);
}
};
Install types:
npm install -D @cloudflare/workers-types
For detailed information, consult:
references/runtime-apis.md
- Complete Workers runtime API documentation with examplesreferences/bindings-guide.md
- All binding types with configuration and usage patterns
Working examples in
examples/
:
fetch-handler-patterns.js
- Common routing and request handling patternserror-handling.js
- Comprehensive error handling strategies
For the latest documentation:
Use the cloudflare-docs-specialist agent to search documentation and fetch the latest information.