Skip to main content
This guide walks through creating a sandbox, waiting for it to become ready, and tearing it down. By default everything runs on the hosted platform at app.amika.dev.

​
Prerequisites

  • An Amika account and an API token. The CLI uses amika auth login; the SDK expects you to source a token via accessToken or tokenSource.
  • (For the --git flow) Git installed and a repo to mount.
  • For the SDK: npm install @amika/sdk.

​
Step 1 — Create the sandbox

The simplest sandbox uses the coder preset (Ubuntu 24.04 with Node, Python, and the Claude Code / Codex / OpenCode CLIs pre-installed). 
amika sandbox create --name dev-box --preset coder
To clone a repo into the sandbox, pass --git (CLI) or repoUrl (SDK). The CLI’s --git accepts a local path or a git URL; the SDK accepts a URL. 
# Clone the repo in the current directory
amika sandbox create --name dev-box --preset coder --git

# Or clone a remote URL
amika sandbox create --name dev-box --preset coder \
  --git https://github.com/gofixpoint/example-repo.git

​
Step 2 — Wait until it’s ready

A new sandbox starts in state initializing and transitions to active / running / started once it’s reachable. 
# `sandbox create` blocks until the sandbox is ready — no extra step needed.
waitForSandbox throws an AmikaError if the sandbox reaches the failed state. New state values can appear over time — the wait helper treats active, running, and started as ready and anything else (except failed) as still working.

​
Step 3 — Tear it down

When you’re done, delete the sandbox to free its resources. 
amika sandbox delete dev-box --force

​
Next steps

Send a message to an agent

Inject credentials

Sandbox lifecycle

Sandbox configuration reference