openwrangler

A library that implements wrangler's remote functionality, allowing you to use Cloudflare bindings via direct REST API calls instead of local simulation.

Overview

openwrangler provides the same binding interface as wrangler's getPlatformProxy, but instead of using local miniflare/workerd simulation, it directly calls the Cloudflare REST API. This enables you to work with remote Cloudflare resources during development.

Installation

npm install openwrangler
# or
pnpm add openwrangler

Direct Usage

You can use openwrangler directly in your Node.js applications:

import { getBindings } from 'openwrangler'

const bindings = getBindings({
  accountId: 'your-account-id',
  apiToken: 'your-api-token',
})

// Use bindings just like in Cloudflare Workers
await bindings.r2.put('image.png', buffer)
await bindings.kv.get('key')
await bindings.d1.exec('SELECT * FROM users')

Nitro Integration

openwrangler can be seamlessly integrated with Nitro applications using @bino0216/nitro-cloudflare-dev.

Setup

1. Install the Nitro integration package:

npm install npm:@bino0216/nitro-cloudflare-dev

2. Configure remote credentials in your Nitro config:

// nitro.config.ts
export default defineNitroConfig({
  cloudflareDev: {
    remoteCredentials: {
      accountId: 'your-account-id',
      apiToken: 'your-api-token',
    }
  }
})

3. Mark bindings as remote in your wrangler.toml or wrangler.json:

# wrangler.toml
[[r2_buckets]]
binding = "MY_BUCKET"
bucket_name = "my-bucket"
remote = true

[[kv_namespaces]]
binding = "MY_KV"
id = "your-kv-id"
remote = true

[[d1_databases]]
binding = "MY_DB"
database_name = "my-database"
database_id = "your-db-id"
remote = true

Or in JSON format:

{
  "r2_buckets": [
    {
      "binding": "MY_BUCKET",
      "bucket_name": "my-bucket",
      "remote": true
    }
  ],
  "kv_namespaces": [
    {
      "binding": "MY_KV",
      "id": "your-kv-id",
      "remote": true
    }
  ],
  "d1_databases": [
    {
      "binding": "MY_DB",
      "database_name": "my-database",
      "database_id": "your-db-id",
      "remote": true
    }
  ]
}

When bindings have remote = true, Nitro will use openwrangler's implementation to connect to your actual Cloudflare resources instead of local simulation.

Supported Bindings

• R2 (R2Bucket) - Object storage
• KV (KVNamespace) - Key-value storage
• D1 (D1Database) - SQL database

All types are imported from @cloudflare/workers-types to ensure compatibility with Cloudflare Workers.

Development

pnpm install
pnpm dev      # Run playground
pnpm build    # Build package

Project Structure

• src/index.ts - Package entry point, getBindings() function
• playground/ - Nuxt/Nitro v3 based test environment
• build.config.ts - unbuild configuration

Use Cases

• Development against production data: Work with real Cloudflare resources during development
• Testing: Test against actual Cloudflare services without deployment
• CI/CD: Run integration tests against remote Cloudflare resources
• Hybrid environments: Mix local and remote bindings based on your needs

License

MIT