Getting Started

Let’s create our first worker and run it locally without deploying.

First, create a file called wrangler.toml:

name = "hello" # Name of the worker
main = "./main.js" # The entrypoint of your worker

wrangler.toml serves as a configuration file which specifies what the worker identifies as and what it can do. To get things up and running quickly, we’ll just specify it’s a worker called “hello” with the worker script located at ./main.js.

Then you can write your worker script in JavaScript in the same directory as wrangler.toml:

export default {
  async fetch() {
    return "Hello, World!"
  }
}

For this worker, all it does is return some text (Hello, World!) with the default status code 200 OK. We can run it locally to see if it works as expected.

We can use the serverlessd CLI to start a “one” instance, which is designed for running one singular worker for testing.

serverlessd one ./wrangler.toml

Open up http://localhost:3000/worker/one (or the port you specified). You should see a piece of text appearing on the page:

Hello, World!

You’re good to go! Learn more about serverlessd and how you can leverage serverless runtimes by exploring the documentation or just trying things yourself. Have fun!