Questions about Cloudfield
What runs on your server, what your team can do, and where the boundaries are. Short answers with links to the source guides.
What it is
What is Cloudfield?
Cloudfield is an open-source multiplayer runtime where engineering teams run and jointly steer AI coding agents like Claude Code, Codex and many more on their own servers. Everyone with access to a workspace can watch its live sessions, and teammates take control of a session from each other. Each session works in its own container on your Linux server and commits to its own git branch. An agent can start subsessions on other agents and models for parts of its task, and every session can open a browser the agent drives and the team watches. The quickstart walks through the whole loop.
copy linkWho is Cloudfield for?
Startups and small engineering teams that already work with coding agents such as Claude Code, Codex or one of the many others, and want them on one server where everyone can see and steer each other's sessions. It fits a team that reviews agent work as branches and pull requests and can run a Linux server. It is not built for anyone who wants a framework for writing their own agents, or for anyone who wants a hosted service. Compare the deployment choices.
copy linkIs Cloudfield self-hosted?
Yes. You operate the Linux workspace server and its storage. Connect directly, including over Tailscale, or use the optional edge for sign-in and relay. The publisher-operated edge stores identity and connection metadata, but cannot read end-to-end encrypted session content. Read the privacy policy.
copy linkHow is Cloudfield different from a hosted coding-agent platform?
Cloudfield's current workspace server runs on infrastructure you operate, with agents installed and signed in through your server-side member environment. You control that host, its data and its upkeep. Cloudfield is GPL-3.0 software; what you pay for is your agent provider and your own server. Compare the deployment choices.
copy linkDo I need an account to use Cloudfield?
Direct server access does not require a Cloudfield cloud account. Access through an edge uses GitHub sign-in and an edge account. Server membership, invitations and device approval still depend on the server's access policy; signing in is not permission to join any server. See access and invitations.
copy linkIs there a hosted Cloudfield workspace?
There is no hosted Cloudfield workspace today: the protocol reserves a kind for one, but no hosted feature exists. You operate the workspace server yourself. The existing optional edge is a sign-in and relay service, not a hosted workspace: it does not run your agents or store your workspace. Read how an edge works.
copy linkIs Cloudfield open source?
Yes. Cloudfield is published by 3xDevOps under GPL-3.0. Bundled dependencies retain their own licenses. Read the source and license.
copy linkWhere is the source code?
On GitHub, at github.com/3xDevOps/cloudfield. Cloudfield is open source under GPL-3.0, so you can read every line of code and run it on your own servers.
copy linkAgents and sessions
Can I run Claude Code on a remote server with Cloudfield?
Yes. Claude Code is a shipped agent. Set it up from the dashboard's Agents page, which installs it in your environment on the server and opens that terminal for its login, then pick Claude Code when you launch a session. It runs in a container on the server, not on your laptop. See the agents guide.
copy linkDoes Cloudfield support OpenAI Codex?
Yes. Codex is a shipped agent. Set it up from the Agents page, or install the Codex CLI in your environment terminal yourself, and sign in there with its device-code login. Then pick Codex for a session. For an API key instead of a login, run codex login --with-api-key there; Codex ignores OPENAI_API_KEY. See the agents guide.
copy linkDoes Cloudfield support OpenCode?
Yes. OpenCode is a shipped agent. Set it up from the Agents page, or install it in your environment terminal yourself, then sign in with its own login command there. Its login stays in your server-side member home and reaches every session you launch. See the agents guide.
copy linkWhich AI coding agents does Cloudfield support?
Cloudfield ships harnesses for many agent CLIs, like Claude Code, Codex, Gemini CLI and GitHub Copilot CLI. Members can register their own launch definitions for other CLIs, and administrators can pin definitions for everyone with inline JSON through a server flag or environment variable, not a JSON file path. Registering an agent does not install it or sign it in. The agents guide lists every one and what it needs.
copy linkDoes Cloudfield install my coding agent for me?
When you ask it to. Set up on the dashboard's Agents page runs the vendor's install command in your environment, your persistent container on the server, then opens that terminal with the agent's login command typed for you to finish. You can also install it yourself there. When a session launches, Cloudfield updates a shipped agent installed in your home in the background, for the next launch, except OpenCode, which stays at the version you installed. The Agents page checks for the executable and a login file; the agent verifies its login when it starts. See the agents guide.
copy linkHow do agent credentials work?
Native vendor logins persist in your server-side member home, mounted read-write into your sessions; a session on your shared account gets only the agent's login file, or one agent CLI's whole agent directory. Cloudfield does not synchronize those credential files back to clients. For supported Claude Code and Codex subscription quota, the server reads native credentials and calls the vendor's HTTPS usage endpoint. Container processes can access credentials in their mounted home. See the security posture.
copy linkDo agents keep running when I close my laptop?
Yes, while the server and session remain up. The server supervises the container and owns the agent's terminal, so closing a client or detaching does not stop the session. Reconnect from the CLI or dashboard to continue watching or steering it. See how a session works.
copy linkHow does Cloudfield run coding agents?
Each session gets its own container, git worktree and branch on the server. It starts from the launcher's saved environment image, or the standard Ubuntu image, with the launcher's persistent member home mounted alongside the checkout, also on another member's shared account. The home persists across sessions; it is not a separate private home for each session. Agents run with their own permission prompts off by default, so a session does not stop to ask before each command; sudo is the exception, and at the Default permission tier it waits for you to allow it. See the environments guide.
copy linkDoes Cloudfield use Docker?
Yes. Docker is required on the Linux server, and sessions and environment terminals execute in containers. This is a container boundary, not an additional agent sandbox: agents run as an unprivileged user, with sudo waiting for your approval at the Default permission tier, and can read and change whatever is mounted into their container. With the optional Sysbox runtime, each session except an Isolated one gets its own Docker daemon, so an agent can build and run containers without reaching the host's. See the install guide's requirements.
copy linkWhat happens when a session finishes?
When an interactive agent reports that it is done, the session reads Finished: review and stays live, so you can read the changes and send follow-ups. Mark complete closes it, and Not yet leaves it open. An agent exiting does not close an interactive session either: it drops to a login shell until someone closes it. Closing commits and publishes the latest work to the session branch and keeps the container, seven days by default, so the session can be relaunched. A Background session commits, publishes and finishes by itself when the agent exits. Read the completion and close workflow.
copy linkDoes Cloudfield merge code automatically?
No. Publish pushes the session branch without force and can open a pull request into the base branch, which you review and merge on GitHub as usual. Pull fetches the session branch into your linked clone without merging it. Subsessions do not change that: a subsession works from its parent's last commit and its result goes back to the parent session, not into your base branch. A session across several repos gets a pull request per repo, with no cross-repo merge. For a workspace mirrored from an upstream repository, a push to the base branch is rejected: the base changes only through upstream. Read how a result is reviewed and published.
copy linkCan one agent direct other agents?
Yes, with subsessions. An agent can start a subsession: a session with a fresh context on any agent and model you have installed, for example Codex reviewing what Claude Code wrote. It appears under its parent in the dashboard and reports its result back to the parent, and you can open, message, take over or stop it like any session. You can turn subsessions off or choose which agents they may use, and a parent has at most eight busy at once by default. A subsession cannot start subsessions of its own. Apart from that, sessions in the same workspace can find each other and exchange messages, and the conflict radar flags sessions changing the same files; neither locks anything or grants control. Read about subsessions.
copy linkDoes each session get a browser?
Yes. A live session can open its own headless Chromium, which Cloudfield starts in a separate container beside the session and drives with Playwright on the server. It shares the session's network, so localhost is the app the agent started. The agent drives it with shell commands Cloudfield puts in every session. Anyone who can steer the session sees the same page in its Browser tab, can take over, and can annotate an element or pick console errors to send to the agent. Nothing starts until someone opens a page; it then reserves one CPU and can use up to 1 GiB of memory. Read about the Browser view.
copy linkTeams
What does multiplayer mean in Cloudfield?
Everyone with access to a workspace can open its live sessions and watch the agent work, and collaborators can steer them. One person controls a session's agent terminal at a time. Take control is immediate when nobody holds it; to take over from a teammate, hold it for three seconds, and they can accept or deny, with control passing to you after seven seconds if they do neither. A message to the agent from anyone but the controller waits 45 seconds for the controller's decision. Everyone who messages a session is credited as a co-author on the commits Cloudfield makes for it. See multiplayer in practice.
copy linkCan teammates watch or steer a session?
Yes. Multiple people can watch a live session's terminal or Chat view at once. Collaborators can steer: take control and type to the agent, or send it a message. Viewers stay read-only. Protecting a session limits messaging, control and killing it to its owner and admins. Notes in a session's Details panel are for people; the agent never sees them. Sending the agent a message or typing into its terminal credits you as a co-author; typing in a separate shell in the session's container does not. See the teams guide.
copy linkWhat does the Map view show?
Map is one of the Board's four layouts, beside Board, List and Lanes. It shows every session on its own, subsessions and finished sessions included, grouped by the person who owns it, with each subsession hanging under its parent. A parent's fold button shows its subsession count, and a teammate's subsession under your session shows its owner's avatar. Read the Board guide.
copy linkCan I cap what a team spends on agents?
A workspace budget can warn before a hard cap and refuse new sessions at the cap unless an override is enabled. It does not stop sessions already working. Displayed spend is accumulated usage, not a forecast; agents that report no usage make the total a floor, not an exact bill. Deleting a session does not erase its counted cost. See the teams guide.
copy linkInstall, data and security
Which operating systems does Cloudfield support?
The server supports Linux on amd64 and arm64. CLI clients support Linux, macOS and Windows. The CLI installs the optional desktop app: it builds the app on your computer on Linux and macOS, and downloads the release's prebuilt app on Windows. Every release includes an Android app, which reaches a server over a tailnet or through the optional edge. See the install guide.
copy linkDoes the Cloudfield server run on macOS or Windows?
No. macOS and Windows are client platforms: you run the CLI there and point it at a Linux server. The install script refuses to put the server on anything else. See the install guide.
copy linkCan I run Cloudfield on an old PC or a home server?
That is the intended setup: a home server, a spare PC, or an old laptop repurposed to stay on all day. The server needs Linux on amd64 or arm64, plus Docker and git, with at least 2.5 GiB of available memory and 6 GiB of free disk. The default limits of up to 8 CPU cores and 8 GiB of memory per session are ceilings, not reservations, and you can lower them. A session's browser, once opened, reserves another CPU and can use up to 1 GiB. See the install guide's requirements.
copy linkCan I use Tailscale with Cloudfield?
Yes. Tailscale is a direct connection option, separate from the optional edge. Tailnet identity can be used when linking the server, and an optional web port serves the server dashboard over HTTPS on its tailnet addresses. Access still follows the server's membership and approval rules. See the networking guide.
copy linkHow do I install Cloudfield?
Choose a Linux workspace server and a client for your computer. The install guide covers the shell installer, Windows PowerShell installer, release checksums and server service setup. Install Docker and git on the server, link your client, then set up your chosen agent and sign in to it in your member environment. See the install guide.
copy linkDoes Cloudfield host my code?
Your workspace repository and session data live on the server you operate. The optional edge relays encrypted connections; it is not code hosting or workspace storage. Model providers can still receive prompts and code through the agent CLIs you run, and connecting an upstream repository adds that destination to your workflow. Read the security boundary.
copy linkWhere does my code and data go?
Agents send prompts and code to their chosen model providers. Cloudfield also uses external services for releases, container images, supported vendor quota reads and, if you turn them on, browser push notifications. If you use the publisher-operated edge, it stores identity and connection metadata, but cannot read end-to-end encrypted terminal, agent, file or dashboard content. Self-hosting does not mean nothing leaves the machine. Read the privacy policy, including edge data.
copy link