About xShellz

xShellz builds the infrastructure teams rely on to run AI where the work actually happens: Agent Shell, a sandboxed cloud box you can hand an AI agent root on; turborg, an always-on AI layer for the channels your team lives in; and borg, the same AI as a coding agent in the terminal. We have operated always-on shell infrastructure since 2009, and today more than 180,000 developers and teams build on it. The intelligence is the new part. The platform underneath has kept people online for well over a decade.

What we do

We run the hard, critical parts of production so your team does not have to. No box to provision, no sandbox to harden, no uptime to babysit. You hand an agent a machine, put AI in your channels, or point it at your repo, and it just works. The infrastructure, the isolation and the AI layer are all ours to keep fast, secure and accountable. Want the detail? See Agent Shell, turborg or compare plans.

What we believe

• Free should be genuinely useful. The free tier is a real box and a real AI budget, not a teaser. When your team outgrows it, the paid plans raise the limits and keep everything always-on.
• Isolation is not optional. Every Agent Shell runs under gVisor so even root inside cannot reach the host, with per-box egress and resource caps. Safe by construction, not by policy.
• Honest pricing. Flat plans you can read top to bottom, no setup fees, no per-seat surprises, cancel anytime.
• Built for the long haul. We have run this infrastructure since 2009. The AI sits on a foundation that is already proven at scale, not a weekend project.
• Open where it counts. The turborg core is open source on GitHub, and we run the managed, always-on platform so your team does not have to.

Built for teams and enterprises

Security is the product, not an afterthought. Agent Shell boxes run under gVisor, a userspace kernel between the box and the host, so even root inside cannot reach the host kernel; container root maps to an unprivileged host user, and hard caps bound memory, CPU, processes, disk and network egress. Sessions use SASL authentication and TLS by default, credentials are encrypted at rest, accounts support multi-factor authentication, and admins can review and revoke signed-in devices at any time. We operate transparently, cooperate on abuse, and comply with valid legal process, as set out in our Acceptable Use Policy and Privacy Policy.

The company

xShellz is an established technology company that has operated continuously since 2009, growing from managed shells into the platform teams reach for when they put AI to work. Longevity is the point: when you build on xShellz you build on a company that has kept its services online for more than a decade and intends to for the next one. Our ambition is to be the default place a developer runs an AI agent, and we build like it.

We are remote-first, with our team working across the United States and the European Union, close to the customers we serve on both sides of the Atlantic. We run the entire stack ourselves, from the always-on infrastructure to the sandbox to the AI layer on top, so the experience stays fast, secure and accountable to one team rather than a chain of vendors.

We are GDPR compliant. If you ask us to remove your data, we honor that right to erasure and delete your personal data from our active systems. Payments are handled by Paddle, our Merchant of Record, so billing is processed securely by an established global provider. See our Privacy Policy for the detail.

Operating since 2009 · 180,000+ developers and teams · Remote across the US & EU · gVisor-isolated · GDPR compliant

Get in touch

Questions, feedback or partnership ideas are always welcome. Reach our team through support, find us on IRC, or open the app. We read everything.