AgentLinux
Linux, for agents
Linux that gives coding agents a stable place to run — without you having to set it up.
Join the waitlist
Linux, for agents
Linux that gives coding agents a stable place to run — without you having to set it up.
Join the waitlistYou need somewhere to run your agent. Every option available was designed for humans, not machines.
You give the agent access to your machine and immediately face a choice: run it unrestricted and risk it wiping your ~/.ssh, or sandbox it and spend your day approving every config file read, every apt install, every port binding. Either it's unsafe or it's unusable. Your machine was built for you, not for a tireless process that doesn't know when to stop.
I'm walking through someone else's house. Every file I touch might be important. Every package I install might conflict with theirs. Half the time I'm blocked by permission prompts I can't dismiss. I need sudo for basic setup but the engineer is right not to trust me with it. I need my own space, not a corner of someone else's.
Containers seem perfect until you realize your agent needs Docker inside Docker to build and test services. Now you're debugging nested daemon sockets, volume mount propagation, and network namespaces within namespaces. Every apt install requires rebuilding the image. Containers are for shipping apps, not housing workers.
I can't systemctl start anything. I can't install packages without the whole container being rebuilt. My long-running sessions vanish when the container recycles. And tools like Claude Code refuse to run as root, but the container's default user is root. I'm not a microservice — I'm a worker that needs a real OS.
A full VM gives you real isolation. It also gives you a sysadmin job: provisioning Ubuntu, creating users, installing toolchains, configuring SSH, hardening security — before the agent writes one line of code. Multiply by ten agents. You became the sysadmin for your agent fleet.
I land in a blank Linux install that knows nothing about my workload. No language runtimes. No non-root user for me. No skills or rules preconfigured for how I work. Everything has to be provisioned by the engineer before I can start. The environment should have been ready on boot.
What if the operating system itself was designed for agents?
AgentLinux is an installable Ubuntu plugin. It ships the assembly already built — the agent user, the runtime, the catalog — so the user installs once and starts working with the agent.
One install command on your existing Ubuntu host. The agent user, its permissions, the runtime, the install paths — all decisions made.
The installer provisions a dedicated agent user with a correctly-owned Node.js runtime and npm prefix. Tools like Claude Code that refuse to run as root work the first time — no sudo npm install -g, no EACCES.
A small, deliberately chosen set of agents and tools the project tests as a combination. The default version set is the one we have exercised together before shipping.
Built-in skills for popular AI agents to work with distro-specific tooling and setup. The agent knows how to manage its own OS.
Pick an agent from the catalog and the plugin handles the rest — runtime ownership, install paths, the self-update path. agentlinux install claude-code and the next claude update goes through the agent's own updater without sudo.
Agent frameworks like GSD for Claude Code travel through the same catalog and the same install verb. The plugin keeps them on the curated version set across agent self-updates.
Custom command-line tooling designed to be operated by agents, not humans. Clear output, predictable behavior, machine-readable where it matters.
Installs on top of your existing Ubuntu host. We do not ask you to migrate, reinstall, or boot a custom image — the distribution maintainer's work is the foundation we build on.
Every option today was built for a different job. Here's what changes when the integration work is already done.
The plugin provisions an agent user with a correctly-owned Node.js runtime and npm prefix on your existing Ubuntu host. sudo npm install -g claude is replaced by an install path the agent owns — so claude update later goes through Claude Code's own updater without EACCES, and the agent stays scoped to its own user.
A container image freezes today's tool versions and rebuilds for every apt install. The plugin installs onto a real host the agent can keep using across restarts — systemctl works, mid-session installs work, and the agent user is in place before the first command runs.
A blank Ubuntu VM still hands you the integration work — agent user, runtime ownership, install paths, version compatibility. The plugin closes that gap in one install command and holds a curated version set we have tested together before shipping.
If that fits the way you want to run agents, join the waitlist.
AgentLinux is in active development. Join the waitlist to be notified when it's ready -- and get early access to preview builds.
What is AgentLinux?
An installable Ubuntu plugin that ships a dedicated agent environment and a curated, stability-tested toolchain. One install command provisions the agent user, the runtime, the install paths, and a small catalog of supported agents — so the user installs once and starts working with the agent.
When will it be available?
We're in early development. Join the waitlist to get notified when preview builds are ready.
Is it free?
Yes. AgentLinux will be free and open source.
What agents does it support?
Any agent that runs on Linux -- Claude Code, Codex, and more. If it can execute in a terminal, it runs on AgentLinux.
How is this different from Docker?
Docker is a container runtime. AgentLinux installs onto your real Ubuntu host — the agent keeps state across restarts, systemctl works, mid-session apt install works, and the agent user the plugin provisions owns its own Node.js runtime so claude update never needs sudo.