# Contents

Dockerfile Support

Most teams already have a Dockerfile. Now agents can use them too.

Daytona now supports building Snapshots directly from a Dockerfile. Just provide the file, add a build context if needed, and Daytona will take care of the rest. A ready-to-run environment, built and sandboxed with no human intervention required.

From Dockerfile to Running Sandbox

If your repo already has a Dockerfile, agents can now use it to get the same environment a human would. No need to re-specify packages or runtime configuration. Just point to the Dockerfile and go.

1image = Image.from_dockerfile("Dockerfile", context_dir=".")
2sandbox = daytona.create(CreateSandboxParams(image=image))

Once built, the Snapshot can be reused to spin up new Sandboxes instantly. Whether you're coding, testing, or debugging, the agent gets the same setup your team already uses.

Perfect for Task-Based Isolation

This is especially useful when agents need to work across projects with different environments. Each repo gets its own Dockerfile. Each task gets its own sandbox.

  • One task, one Dockerfile, one sandbox.

  • No guesswork.

  • No mismatch between dev and agent environments.

Works With What You Already Have

No need to rewrite or adapt your existing dev setup. Daytona works with your Dockerfiles out of the box. That means your agents can spin up compatible environments using the same infra your team relies on.

Dockerfile support is now live in the Daytona SDK.

If your agent sees a Dockerfile, it knows what to do.

Tags::
  • Dockerfile support
  • AI agents
  • Daytona SDK
  • environment automation
  • agent workflows
  • sandbox environments
  • devtools