Your repositories,
already set up as projects
Connect GitHub or GitLab once, see the repositories you actually have access to, tick the ones you want and clone them in a single pass. Each repository lands as an AgentsRoom project with its path already set, ready to receive agents.
Self-hosted GitLab instances are supported: point AgentsRoom at your own host and everything works the same way.
Connect, pick the repositories, clone. Each one becomes a project with its own agents.
Setting up a new machine, or trying AgentsRoom for the first time, used to mean the same tedious loop: open a terminal, remember the clone URL, clone, come back, create a project, browse to the folder. Repeat for every repository you actually work on.
The import flow replaces that loop. You connect your git host once, AgentsRoom lists the repositories you are a member of, and you select as many as you want. They are cloned in batch and registered as projects in one action.
This is a desktop-only, local flow. There is no server round-trip, no OAuth app to authorize, no copy of your repository list anywhere but on your machine. The token stays encrypted on disk with the operating system's own secure storage, and it never crosses into the interface layer.
What the import actually does
Two hosts, same flow: GitHub and GitLab, including self-hosted GitLab instances behind your own domain.
Two ways to connect: paste a personal access token, or let AgentsRoom reuse the gh or glab CLI you are already signed in to, with a single click and no token to create.
Batch clone: tick several repositories, clone them all in one pass, and get one AgentsRoom project per repository with its path already configured.
Read-only by design: the token scope you are asked for is the minimum needed to list and clone. AgentsRoom never needs write access to import.
How it works
From a fresh install to a room full of projects, without opening a terminal.
Connect your git host
From the project opening screen, pick GitHub or GitLab. Paste a personal access token, or use the one-click option if you already have the gh or glab CLI authenticated on this machine. For GitLab, enter your own instance host if you are not on gitlab.com.
Browse what you actually have access to
AgentsRoom lists the repositories you own or are a member of, with their namespace, visibility, star count and last activity, so you can find the right one without leaving the app.
Select several at once
Tick every repository you want on this machine. There is no reason to import them one at a time when you are setting up a new laptop.
Clone in one pass
The clones run with a transient credential passed only on the command line for that operation. Nothing is written into the repository's git config, so the cloned repositories carry no trace of your token.
Start assigning agents
Each cloned repository is registered as a project, with its folder already set. Add your agents, pick a provider per agent, and start working. Project settings, memory and backlog follow the project from there.
What happens after the import
A project is the unit everything else hangs off. These are the three surfaces you will meet first.
Every imported repository becomes a project in the sidebar, with its own agents running in parallel and its own live status.
Learn moreSwitch the working branch from the app, review what an agent changed file by file, and keep parallel work isolated with git worktrees.
Learn moreKnowledge about a codebase that your agents write and read as they work, keyed to the project so it follows the repository.
Learn moreHow the credentials are handled
FAQ
Which git hosts are supported?
GitHub and GitLab, including self-hosted GitLab instances: enter your own host and the listing, token deep link and cloning all follow it. GitHub Enterprise is not supported yet, and Bitbucket and Gitea are not implemented.
Do I have to create a personal access token?
No, if you already use the official CLI. AgentsRoom can reuse the credentials of a gh or glab session that is already authenticated on the machine, in one click. If you prefer a token, the app links straight to the creation page with the right scope prefilled.
What token scope does it need?
The minimum for listing and cloning: the repository read scope on each host. AgentsRoom does not need write access to import a repository. If you later want an agent to push, that is your normal git configuration, not this token.
Where is my token stored?
In a file inside the AgentsRoom home directory on your machine, encrypted with the operating system's secure storage and restricted to your user. It is never sent to any AgentsRoom server, and the interface layer only ever sees whether a connection exists, who it belongs to and which host it points at.
Can I import several repositories at once?
Yes, that is the point of the flow. Tick as many as you want and they are cloned in a single batch, each one registered as its own project.
Does the clone carry my credentials?
No. Authentication is supplied as a transient header on the clone command itself and is never persisted into the repository's git config. The cloned repository is a normal clone.
Can I still open a repository I already cloned by hand?
Yes. The import is a shortcut, not a requirement. Opening an existing folder as a project works exactly as before, and a repository imported this way is indistinguishable from one you cloned yourself.
Does this work with a monorepo, or with several projects in one repository?
The import registers one project per repository, pointing at the repository root. If you want a project scoped to a sub-folder of a monorepo, open that folder as a project manually after the clone.
You may also like
Multi-project · Multi-agent · Parallel
AgentsRoom runs multiple Claude Code, Codex CLI, Antigravity CLI, OpenCode, Aider, Grok Build, Mistral Vibe and Kimi Code AI agents in parallel across multiple projects, all in a single window. Color-coded roles, live status, per-project counters, native desktop and mobile notifications. The visual alternative to 12 terminal tabs and the chaos of Claude Code sessions scattered across iTerm2, Warp and tmux.
Project Memory
Give your AI coding agents a long-term memory. Decisions, architecture and pitfalls persist across sessions and are shared with everyone on the project. Free, unlimited.
Dev terminals
Run the build, the tests and the dev server of an imported project in the same window as its agents, with a bell when a long command finishes.
SSH connections
Save the machines you code on and run agents straight on a remote server, without leaving the same cockpit.
Commit Context · Agent context handoff
Commit Context attaches the full agent conversation behind each commit as an unlisted gist, linked from the commit message. Recover the reasoning behind any change and hand the context to the next AI agent. Works with Claude Code, Codex, Antigravity CLI, OpenCode, Aider, Grok Build, Mistral Vibe and Kimi Code.
Remote Fleet
Run Claude, Codex and Antigravity coding agents across every Mac you own, and invite teammates to collaborate on the same agents in real time. Office machine, home machine, build server, shared team projects : all in a single unified, end-to-end encrypted view.
Set up a new machine in one pass
Connect, tick the repositories that matter, clone. Free to download on macOS, Windows and Linux.
Companion app: monitor your agents on the go
Bring your own: Claude, Codex, Antigravity CLI, or other AI provider.
Push bugs and requests straight to your public backlog.
A glimpse of AgentsRoom in action.