Search or ask Docker

Get started with Docker Agentic Platform

Review a sample project with a coding agent or build and preview it from a shell in a cloud sandbox, without setting up a local development environment.

View Markdown

Use a cloud sandbox to review a project with a coding agent or build and run an application from a shell, without setting up the project on your computer.

Both paths introduce the same cloud sandbox lifecycle: launch an environment, run a task, save work, and pause or delete it. Complete either path in your browser, then use the guides for sharing, remote development, or moving local work to the cloud.

Before you begin

You need a personal Docker account with an active Docker Agentic Platform subscription.

Both workflows use a public sample project, so you don't need a GitHub token. Review the compute price and timer before launching. Choose Stop as the timeout action to keep files. Compute is billed while the sandbox runs; closing the browser doesn't stop it.

Choose a workflow

  • Use a coding agent to have Claude Code review a sample project and save its findings. You need an Anthropic API key.
  • Use a shell to build and run the sample application and preview it in your browser. You don't need a model provider API key.

Use a coding agent

Add an Anthropic API key under Secrets or enter it when you launch the sandbox. Your model provider bills inference separately from Docker's sandbox compute charges. For other agents and their credentials, see Kits and Secrets.

Launch Claude Code

  1. Open Docker Agentic Platform and select + New sandbox.
  2. Choose the Claude Code kit.
  3. Add your Anthropic API key if it isn't already saved. Don't enter it in the agent's prompt or in a file inside the sandbox.
  4. Keep the defaults and select Run.

On your first launch, the default Open network policy permits outbound access to any host. If you're using a custom policy, make sure it permits GitHub and your model provider. See Network policies.

When the sandbox is ready, its terminal opens with Claude Code.

Give the agent a task

Your local files aren't mounted in a cloud sandbox. Ask Claude Code to clone and review a public sample project:

Clone https://github.com/docker/welcome-to-docker.git into
/home/agent/workspace/project, then review the application.
Read the source, package.json, and Dockerfile. Don't modify the application
or install dependencies.
Write /home/agent/workspace/review.md with:
- A short explanation of what the application does.
- The commands defined in package.json for running and building it.
- An explanation of how the Dockerfile builds and serves the application.
- One small improvement you'd recommend, with the relevant file path.
Base the review on the files you read. Don't claim to have run the application.
Summarize your findings in your response.

Claude Code summarizes the project in the conversation and saves the review at /home/agent/workspace/review.md. You can ask follow-up questions or give it another task in the same session.

Continue to keep your work.

Use a shell

Launch a shell

The Shell kit starts a Bash shell without an agent.

  1. Open Docker Agentic Platform and select + New sandbox.
  2. Open the kit picker and choose shell.
  3. If saved secrets or a GitHub token are selected, deselect them for this walkthrough. You don't need MCP tools either.
  4. In the network policy picker, select Balanced to permit access to GitHub, Docker Hub, and npm. Deselect Open if selected: combining it with Balanced doesn't restrict Open's access. Review the compute price and timer, choose Stop as the timeout action, and select Run.

If you use a custom network policy, make sure it permits GitHub, Docker Hub, and npm. See Network policies.

When the sandbox is ready, its terminal opens with a Bash shell.

Build and run an application

Your local files aren't mounted in a cloud sandbox. In the browser terminal, clone a public sample project:

$ git clone https://github.com/docker/welcome-to-docker.git project
$ cd project

The Shell kit includes Docker Engine. Build the application's container image:

$ docker build -t welcome-to-docker .

The build installs the application's dependencies and compiles it inside the sandbox. When it finishes, start a container and publish its port:

$ docker run -d --name welcome-to-docker -p 3000:3000 welcome-to-docker

Preview the application

  1. Select Connect on the sandbox's detail page.
  2. Under Connect via a Port, enter 3000 and select Open Port.
  3. Open the generated URL in your browser.

You should see the Welcome to Docker application. The application runs in the cloud sandbox, so you can try it without installing Docker or Node.js on your computer.

The generated URL is public, and Docker account authentication doesn't protect it. Use only non-sensitive data. Close the port in Connect when you've finished previewing, then reload the URL to confirm access is withdrawn. For a complete sharing workflow, including checking access from a separate browser session, follow Share an application.

Continue to keep your work.

Keep your work

The cloned project and any files you create are stored in the sandbox. Files don't sync automatically with your computer or a remote repository. They remain available when you pause and resume the sandbox, but are lost when you delete it.

To keep files outside the sandbox, use the CLI to transfer files, or commit and push changes to a repository. Include a GitHub credential when launching a sandbox that needs to push changes.

Pause, keep running, or delete the sandbox

Closing the browser leaves the sandbox running. Commands and background processes can continue until you pause or delete the sandbox, or its timer expires. Compute is billed while the sandbox runs, even when no commands are running.

Return to Sandboxes in the Console and choose what to do next:

  • Pause the sandbox to stop compute charges and keep its files. Confirm that it is paused in the list. To continue later, reopen the sandbox and resume it. Resuming starts a fresh timer with the original duration.
  • Leave it running if the agent still has work to do or to keep the application available. Check the expiration time: the default Stop action stops the sandbox after one hour, including background processes, and keeps its files.
  • Delete it when you no longer need it. Push any source changes you want to keep first. Deletion removes files stored only in the sandbox.

For more detail, see Lifecycle actions.

Sandbox options

For launch settings, see Kits, Secrets, Network policies, and MCP tools. For compute platforms and timer actions, see Sandboxes. Choose these settings before launching: credentials, tools, network policies, and compute size can't be changed afterward.

Next steps

To run a task from your own terminal, follow the cloud CLI walkthrough. For programmatic workflows, see the Sandboxes API quickstart.

Continue with a targeted workflow: