Quick Start — Getting Started

This guide gets you from zero to a running Aerostack project. You’ll install the CLI, create a project, and deploy your first function to the edge.

Install the CLI
Terminal window
curl -fsSL https://get.aerostack.dev | sh
Terminal window
npm install -g @aerostack/cli
Terminal window
pnpm add -g @aerostack/cli
Terminal window
bun add -g @aerostack/cli
Verify it works:
Terminal window
```
aerostack --version
```
Authenticate
Terminal window
```
aerostack login
```
This opens your browser to sign in. Once authenticated, the CLI stores your credentials locally.

Don’t have an account yet? Run aerostack login — it will guide you through registration.
Create a Project
Terminal window
```
mkdir my-project && cd my-project
aerostack init
```
The CLI asks what you need (Database, Auth, AI, etc.) and scaffolds a ready-to-run project with:
- src/index.ts — your function entry point
- aerostack.json — project configuration
- aerostack.toml — Cloudflare Workers configuration
Start Local Development
Terminal window
```
aerostack dev
```
Your function is now running at http://localhost:8788. The local dev server emulates the full Cloudflare edge runtime — including database, cache, and queue bindings.
Terminal window
```
curl http://localhost:8788/health
# {"status":"ok","project":"my-project"}
```

Write Your First Function

Open src/index.ts. This is a fullstack edge function with native access to all platform primitives:

export default {
  async fetch(request: Request, env: Env): Promise<Response> {
    const url = new URL(request.url)

    if (url.pathname === '/api/hello') {
      // Native database — direct binding, no HTTP call
      const result = await env.DB.prepare(
        'SELECT COUNT(*) as count FROM visits'
      ).first()

      // Native cache — same datacenter
      await env.CACHE.put('last-visit', new Date().toISOString())

      return Response.json({
        message: 'Hello from Aerostack!',
        visits: result?.count ?? 0,
        cached_at: await env.CACHE.get('last-visit')
      })
    }

    return Response.json({ status: 'ok' })
  }
}

Deploy to Edge

When you’re ready to go live:
Terminal window
```
aerostack deploy
```
Your function is now running on Cloudflare’s global edge network — 300+ cities worldwide.

What to Build Next

Pick the path that matches what you’re building:

Build a Function — Write fullstack edge functions with native DB, Cache, Queue, AI, and Storage access. The foundation for everything else.
Build a Bot — Deploy an AI bot to Telegram, Discord, WhatsApp, or Slack. Connect MCP tools. Add workflows for complex logic.
Set Up MCP Servers — Host MCPs, proxy existing ones, or install from Hub. Give your team secret-free access with per-user analytics.
Create a Workspace — Compose multiple MCP servers behind one gateway URL. Connect from Claude Desktop, Cursor, or any MCP client.

Using the Dashboard Instead

Everything you can do with the CLI, you can also do through the Admin Dashboard at app.aerostack.dev:

Create and manage projects
Deploy functions, MCP servers, and skills
Build and configure bots with the visual workflow editor
Compose workspaces and manage team access
View analytics, logs, and billing

The dashboard is the recommended starting point for bots and workspaces, while the CLI is best for functions and MCP server development.