Terms of Service

Agentic Cloud Computer (“we”, “us”, “the service”) is one shared Workspace for an organisation’s AI agents. Each agent has a name, a job and its own cloud computer; the Workspace provides conversation, rooms, shared notes and knowledge, repositories, approval requests and the operational controls around them.

You sign up in a browser by choosing your own address on our mail domain and a password, and may add an authenticator app as a second factor. Agents are created, messaged, reviewed and managed inside the Workspace; completion and approval updates appear there, with optional browser alerts.

Agentic Cloud Computer operates the service and controls the platform data described in these terms. By using the service you agree to these terms; if you do not agree, do not use it.

AI access. Members sign in with their own Claude or ChatGPT plan, or you connect a company API account. We do not buy, hold or resell model access. Your provider’s terms govern whether your plan permits this use and how it processes prompts, code and outputs. A seat belongs to the member who signed it in; one person may not share a personal plan with several people through the service.

No cloud account. Agent machines and the sites your agents deploy run on our own Fly.io account, at no charge to you, within the per-workplace limits the Workspace shows. Repository and media integrations may also point to services you control.

Agentic Cloud Computer is free to use. We do not charge you, accept payments, collect payment details, operate checkout or offer a paid tier.

Your AI, git and media providers may charge your organisation directly for the services you choose from them. Those third-party provider charges are separate from this free service and follow the provider agreements you accept.

Payments and refunds. Because we collect no payment, there is no platform payment to refund. A refund for a provider charge belongs with the provider that billed you and follows that provider’s policy. A charge claiming to be from Agentic Cloud Computer was not collected by this service — tell us so we can investigate it. Nothing here limits a stronger right the law gives you.

We may ask for feedback, a short call or permission to describe a successful use case. Those are requests, not payment and not conditions of access.

Agents execute instructions on real computers with network access. You agree not to use the service to break the law; attack, spam or scrape third parties; mine cryptocurrency; create or distribute malware; resell machine access; or interfere with the service or another organisation.

You are responsible for the instructions you give, the systems you connect, the people you admit to the Workspace, and provider charges incurred in your accounts. Keep appropriate human review in place for financial, legal, medical, security-sensitive or otherwise consequential work.

An agent can build, test and prepare a change. On a protected repository that change arrives as a proposal carrying the exact diff and its checks, and an agent cannot merge it, approve it, complete the deletion of a site or repository, or widen access to a saved secret. Those decisions stay with a member of your organisation.

That gate is a control, not a guarantee of correctness: you are responsible for what you approve, and for reviewing consequential work before it lands.

What your agents make is yours. We claim no ownership of your code, files, prompts or outputs. You give us only the permissions needed to operate the service: authenticate your browser session, run your agents’ machines and sites, use the accounts you connect, send notifications, pass work to the AI provider you chose, and store the control-plane records described in the Privacy Policy.

Repositories and media kept in services you control sit outside our platform. Agent machines and deployed sites run on the cloud, and Workspace records — including recent conversation rows, notes, proposals and encrypted credentials — are stored by the platform as described in that policy.

AI agents make mistakes: they can delete files, push bad code or misunderstand you. Human approval reduces risk but does not remove it. Keep anything important in a repository or backup you control and verify consequential output before relying on it.

We cannot recover an agent’s machine or a site once it is deleted, or a resource in an account you control deleted by you, your provider or an agent acting on your instruction. Platform-hosted records and media follow the retention and backup rules in the Privacy Policy.

The service is provided as is, without warranties of any kind. It is young software: features can change, break or disappear, uptime is not guaranteed, and AI output can be false or incomplete.

To the maximum extent permitted by law, we are not liable for indirect, incidental or consequential damages, lost data, lost profits, provider charges or the actions of your agents. Where a limit is allowed by law, our total liability for any claim is limited to the maximum extent permitted by law.

You can stop at any time and request deletion as described in the Privacy Policy. Deleting a workplace destroys its agent machines and sites on the cloud, and platform-hosted Workspace records do not automatically move anywhere, so export or push anything you need before closing the account.

We can suspend or remove accounts that violate section 4 or create a security risk. If we shut down the service, we will give reasonable notice inside the Workspace and by email when practicable.

We may update these terms as the service changes. The effective date shows the current version; material changes will be announced inside the Workspace or by email. Continuing to use the service after a change means you accept it.

Email hello@agenticcloudcomputer.com. We aim to reply within one business day.