AgentsRoom Is Now on Linux

AgentsRoom runs on Linux. You can download it today as an AppImage that works on any modern distribution, or as a .deb package for Debian and Ubuntu. Same app, same features, same visual cockpit for your AI coding agents.

If you have been waiting on a Linux build, this is it. Here is why it took until now, and what you get.

Why we started with macOS, and why we are opening up now

We made a deliberate call at the start: ship one platform, make it excellent, then expand. macOS came first. Building for a single target let us move fast on the hard parts, the parts that have nothing to do with which OS you run: persistent terminal sessions per agent, live status tracking across dozens of agents, end-to-end encrypted sync with the mobile app, the MCP servers that let agents drive the IDE itself.

That bet paid off. The macOS app has been downloaded thousands of times, and hundreds of developers now run AgentsRoom every day to orchestrate Claude, Codex, Antigravity CLI, OpenCode and Aider across their projects. That is the signal we were waiting for. A tool that people actually keep open all day, not one they try once and forget.

So we are doing the obvious next thing: accelerating. The foundation is solid and proven, the audience is real, and a large share of developers running CLI coding agents live on Linux. Keeping them out no longer made sense. Linux is the natural second home for AgentsRoom, and it is here now rather than "someday on the roadmap."

Full parity, not a stripped-down port

This is not a limited preview. The Linux build is the same AgentsRoom you have seen on macOS, feature for feature:

• The multi-project, multi-agent cockpit, with every agent in its own real terminal and a live status dot.
• Dev terminals for your backend, frontend, bundler and database, launchable in one click.
• Native desktop notifications when an agent finishes or needs your input, wired into your Linux notification daemon, with a tray presence so you can step away and still know what is happening.
• Mobile sync over the same end-to-end encrypted relay, so you can monitor and steer your agents from your phone whether your desktop is a Mac or a Linux box.
• The four MCP servers (Backlog, Terminal Commands, Prompt Library, Browser) that turn AgentsRoom into an agent-controllable IDE.

Everything runs locally. Your agents spawn as real CLI processes in your own project folders, exactly as they would in your terminal. No cloud relay for the code, no proxy, nothing leaves your disk.

Installing on Linux

Two options, pick what fits your setup:

• AppImage: one self-contained file. Make it executable and run it, no installation and no root required. Works across distributions.
• .deb: a standard package for Debian, Ubuntu and derivatives, if you prefer managing it through your package manager.

You need a 64-bit distribution and at least one supported CLI installed (Claude, Codex, Antigravity CLI, OpenCode or Aider). That is the same single requirement as on macOS: AgentsRoom is the cockpit, your CLI of choice does the coding.

Grab the build from the download page and you are running in under a minute.

What comes next

Windows is the remaining gap, and it is on the roadmap. Bringing AgentsRoom to Linux moved us a long way toward a fully cross-platform desktop app, and Windows builds on much of the same work.

For now: if you code on Linux and you run more than one agent at a time, AgentsRoom was built for exactly that. Download it, connect your provider, and put your whole fleet on one screen.