The work, documented.
The console is available today. TypeScript and Python SDK source is public and audited; installable releases remain closed until the release authority is complete.
for your agent · paste it into Claude Code, Codex, or the agent you use
Read Cuna's docs: https://getcuna.com/docs My project: [describe it: stack, repo, what you want agents to do] Context, so you answer precisely. Cuna runs coding agents on isolated cloud machines the user controls. In the console at app.getcuna.com the user creates a machine, picks an agent (a clean machine, Claude Code, Codex CLI, or OpenClaw), sizes it (up to 2 vCPU and 4 GiB), opens it in the browser, and hands it the task. The user starts, pauses, stops, and deletes the machine; nothing runs on its own. Every action the machine takes is written to the record, readable as it happens. The console is the available product today. TypeScript and Python SDK source previews are public at github.com/Cuna-Labs/cuna-lib-ts and github.com/Cuna-Labs/cuna-lib-py, but there is no GA registry release, public API, or Cuna CLI yet. The npm package is only a bootstrap reservation and the Python package is not on PyPI. Answer three things, concretely: 1. Which tasks in my project should become agent runs. 2. Which agent to use for each, and how to size and stop the machine. 3. Where to start in the console: which agent, and the first task to hand it.
01
What Cuna is
Cuna runs coding agents on isolated cloud machines you start, size, and stop — and keeps a record of everything they do. Today it all lives in one place: the console.
Give agents the work, never the keys.The work ends. The proof remains.
02
Quickstart
Nothing to install. Open the console and put an agent to work in four steps.
Sign in
If a workspace is available, it is assigned immediately. Otherwise, your place on the waitlist is saved and we'll email you when one is ready.
Create a machine
Pick an agent, size it, and start it — chapter 03.
Open it
Open the machine in your browser and hand it the task.
Read the record
Every action the agent takes is on the record — read it as it happens.
Everything runs from your browser — no CLI, no local setup.
03
Machines
The work happens in a machine: an isolated cloud sandbox with its own files, tools, and clock. You create one, choose the agent, open it, and stop it when the task is done.
clean machineJust the environment — bring your own agent.Claude CodeAnthropic's coding agent, ready in the machine.Codex CLIOpenAI's terminal coding agent, ready in the machine.OpenClawThe open coding agent, ready in the machine.
Size it up to 2 vCPU and 4 GiB; one machine runs at a time for now. Optionally publish a port to reach a service running inside the machine from your browser.
Create & open
The machine starts in seconds; open it in your browser.
Pause, resume
Pause a machine to hold it; resume it when you return.
Stop, start
Stop it to free the compute; start it again anytime.
Delete
Remove the machine for good. The record of what it did remains.
04
The record
Every action on a machine is written to the record as it happens — created, opened, commands run, paused, stopped. It is not a log you scroll away; it is the account of what the agent did. Read it per machine, or across your whole workspace.
created claude-code · 2 vCPU · 4 GiBopened session readyexec npm test · 214 okpaused state heldstopped machine closed
You do not wonder what the agent did. You read it.
05
Honest boundaries
Trust is the product, so the boundary is stated. Cuna gives you isolated machines you start and stop, and a record of every action. It does not verify that the agent's output is correct.
do
- Trust that each machine is isolated — its own files, tools, and clock.
- Trust the record — every action is written as it happens.
- Trust that you start, size, and stop the machine; nothing runs without you.
don't
- Assume an answer is true because the run was clean — read it, test it.
- Assume generated code is free of bugs.
- Assume a machine hides your data from the platform operator.
06
SDK source previewsource open
Cuna's TypeScript and Python SDK repositories are public and have passed their release-quality gates. Both clients are built for the exact Cuna API origin at https://api.getcuna.com; they do not embed the Cuna backend or its credentials.
@cuna_labs/sdkTypeScript ESM client for Node.js 22+. The source is public; npm currently contains only the 0.0.0-bootstrap.0 registry reservation.cuna-sdkPython 3.10+ synchronous and asynchronous clients. The source is public; no PyPI distribution is published yet.