Part 1: Your First Stack

In this part you install Alchemy, create a Stack with a Cloud Storage Bucket, and deploy it to your Google Cloud project.

Prerequisites

• Bun or Node.js 24.11.1+
• A Google Cloud project and a service-account key that can create what you deploy. See Setup.

Create a project

Start with an empty directory and a package.json:

pnpm

mkdir my-app && cd my-app && pnpm init

bun

mkdir my-app && cd my-app && bun init -y

npm

mkdir my-app && cd my-app && npm init -y

yarn

mkdir my-app && cd my-app && yarn init -y

Install dependencies

Install Alchemy, the GCP provider, and Effect:

pnpm

pnpm add https://pkg.alchemy.run/alchemy/a870ed4 https://pkg.alchemy.run/@alchemy.run/gcp/a870ed4 effect @effect/platform-bun @effect/platform-node

bun

bun add https://pkg.alchemy.run/alchemy/a870ed4 https://pkg.alchemy.run/@alchemy.run/gcp/a870ed4 effect @effect/platform-bun @effect/platform-node

npm

npm install https://pkg.alchemy.run/alchemy/a870ed4 https://pkg.alchemy.run/@alchemy.run/gcp/a870ed4 effect @effect/platform-bun @effect/platform-node

yarn

yarn add https://pkg.alchemy.run/alchemy/a870ed4 https://pkg.alchemy.run/@alchemy.run/gcp/a870ed4 effect @effect/platform-bun @effect/platform-node

Create the Stack

Every Alchemy program starts with a Stack. Create alchemy.run.ts:

alchemy.run.ts

import * as Alchemy from "alchemy";
import * as GCP from "@alchemy.run/gcp";
import * as Effect from "effect/Effect";

export default Alchemy.Stack(
  "MyApp",
  {
    providers: GCP.providers(),
    state: GCP.state(),
  },
  Effect.gen(function* () {
    // resources go here next
  }),
);

The first argument, "MyApp", names the Stack. Alchemy uses it as a prefix when it generates names for your resources.

GCP.providers() registers every GCP resource.

Where state lives

state: GCP.state() keeps deploy state in a Cloud Storage bucket in your project, alchemy-state-{project}, created on the first deploy. Because state lives in your project and not on your machine, the same setup works on your laptop and in CI.

Other state stores

For purely local iteration, use Alchemy.localState() instead. It writes state under .alchemy/ next to your code. See State store.

Connect your project

Point Alchemy at your project once:

bun alchemy profile edit --add GCP

Choose Service account JSON, pick the key file, and pick a default region. The choice is saved to your default profile.

Add a Bucket

Declare a Cloud Storage Bucket inside the Stack:

alchemy.run.ts

import * as Alchemy from "alchemy";
import * as GCP from "@alchemy.run/gcp";
import * as Effect from "effect/Effect";

export default Alchemy.Stack(
  "MyApp",
  {
    providers: GCP.providers(),
    state: GCP.state(),
  },
  Effect.gen(function* () {
    const bucket = yield* GCP.Storage.Bucket("Bucket", {});
  }),
);

Bucket is the logical id. Alchemy generates a globally unique bucket name from the Stack, stage, and logical id.

Return Stack outputs

Return the values you want to see after a deploy:

Effect.gen(function* () {
  const bucket = yield* GCP.Storage.Bucket("Bucket", {});

  return {
    bucketName: bucket.bucketName,
  };
}),

bucket.bucketName is an Output. It resolves during the deploy and prints with the rest of the outputs.

Deploy

pnpm

pnpm alchemy deploy

bun

bun alchemy deploy

npm

npx alchemy deploy

yarn

yarn alchemy deploy

Plan: 1 to create
+ Bucket (GCP.Storage.Bucket)

Proceed?
◉ Yes ○ No
✓ Bucket (GCP.Storage.Bucket) created
{
  bucketName: "myapp-live-sam-bucket-a1b2c3d4",
}

Alchemy shows the plan, asks for confirmation, creates the Bucket, and prints the outputs.

Verify it worked

The Bucket shows up under Cloud Storage → Buckets in the console, or with:

gcloud storage ls

Deploy again. Nothing changed, so there is nothing to do:

Plan: no changes

{
  bucketName: "myapp-live-sam-bucket-a1b2c3d4",
}

Destroy (optional)

destroy deletes everything the Stack created:

bun alchemy destroy

Deploy again afterwards if you try it. Part 2 builds on the Bucket.

Recap

You now have:

• An alchemy.run.ts with a Stack and a Cloud Storage Bucket
• A live Bucket in your Google Cloud project
• Stack outputs showing the Bucket name

In Part 2, you put an HTTP service on Cloud Run that reads and writes objects in this Bucket.