Skip to content
Sign up

A development machine that never turns off.

Give coding agents a persistent workspace to build, run and test real software.

Sign up See how it works

Free and open source. Bring your own server.

A daemon working at a laptop on top of a black development machine that shows a terminal and a live preview, wired to four chips for Claude Code, Codex, Cursor and OpenCode.

Always on

Agents, apps and databases keep running 24/7.

Safe by isolation

Agents only see what's inside their workspace, not your laptop or local data.

Access anywhere

Keep building from browser, IDE, CLI or phone.

Run real software,
not just agents.

Your agent works inside a complete development environment that stays there between sessions.

Install once

Docker, databases, dependencies, browsers, tools.

Keep it running

Frontend, backend, workers and services stay alive.

Access anywhere

Browser, IDE, CLI, SSH or phone.

A daemon at a laptop between a terminal running npm run dev and a browser window showing My App live.
Postgres
Redis
Docker
App running

Do you actually know how your coding agents are working?

Does work stop when laptops close at 4pm?

Can an agent access data it should never see?

Are developers using the right models, or just burning money?

Can you trust every skill and MCP your agent installs?

Once agents become part of engineering, their environment becomes infrastructure.

A closed laptop asleep, and a lime arrow to a cloud where a daemon keeps working on a small stack of servers.

Laptop

Always on
in the cloud

A developer laptop spilling API keys, personal files and credentials, and a lime arrow to a wireframe cube where a daemon works with just its folders and database.

Developer laptop
(too much access)

Isolated daemon workspace
(just what it needs)

Give agents their own place to work.

Instead of letting autonomous software inherit everything on a developer laptop, give each agent an isolated workspace.

Isolate

Agents only see what belongs in their workspace.

Control

Decide which repositories, services and networks they can reach.

Persist

Keep the machine, software and project state between sessions.

Bring the agents you already use.

Claude Code, Codex, Cursor, OpenCode and more.

Switch agents without rebuilding your environment.

Claude Code
Codex
Cursor
OpenCode

One workspace.
Any agent.

Set it up once. Keep building.

Step 01: Create

Choose a workspace and connect your project.

Create

Step 02: Configure

Install your stack and sign into your agents.

Configure

Step 03: Keep going

Come back tomorrow, from another computer or from your phone.

Keep going
A daemon mascot waving beside a workspace box.

Try working without depending on your laptop.

Free and open source. Bring your own server.

Sign up

Persistent workspace · Your agents · Your full dev stack