A day with Foreman
Your AI pair programming workflow, from 9:00 to ship.
Foreman is a free, open-source desktop app for macOS and Windows that runs AI coding agents in parallel. Claude Code, Codex, Cursor and OpenCode each work in their own git worktree on their own branch, while you review diffs, answer questions and open pull requests from one window.
macOS & Windows · Apache-2.0 · No account · Your own agent logins
Start the dayOne developer, one repository called atlas, one Tuesday.
Coffee
Three issues. Three agents. Three worktrees.
Three issues came in overnight, so you open three workspaces in atlas. Claude Code takes the login redirect, Codex the rate limiter, Cursor the launch page. Each workspace is its own git worktree on its own branch: nobody edits the same files, and your main checkout stays clean.
- lisbonClaude Code
fix-login-redirect - osakaCodex
add-rate-limit - kyotoCursor
launch-2.4
One worktree per agent

lisbon, Claude Code's workspace: the chat in the middle, its three changed files on the right. - lisbonClaude Code
Tiles
Watch them all. Answer the one that asks.
Tiles puts your workspaces side by side. Most are busy; one has stopped to ask which day the release goes out. You answer in its tile and it carries on. Nobody else had to wait for you.
Built and load-tested for 10 to 12 agents running at once.
Watch them all at once

Tiles: four workspaces, one question waiting for an answer. Approval
A push that says where it's going.
kyoto wants to push. The approval breaks the command into its parts and marks the one that matters: one commit, to origin, as you. It leaves this computer and can't be undone. You read it. You approve.
▲ Leaves this computer And if you reach for Always allow, it says exactly what the rule would add first.
Approvals that spell it out

The approval in kyoto's tile: what the push does, and that it leaves this computer. Lunch
Approve from your phone, between bites.
osaka finishes the rate limiter while you're out and asks to push. The phone view updates live, so the request is already waiting. Go on, tap Approve.
Pair a phone once by matching a six-digit code, and choose what it may do: Watch, or Operate.
Run it from anywhere
ForemanLiveNeeds you 1
osaka Codex · Push
Push
add-rate-limitto originLeaves this computer
Pushes 2 commits to origin (github.com/acme/atlas) as @you. It can't be undone.
Approved. osaka pushed 2 commits.
Held. The commits stay on this computer.
Paired · Operate · same sign-in as your desk
An illustration of the phone view with one push waiting for you. Loop
Red tests? Let a flow loop until green.
The session tests are red and you'd rather not babysit them. You start fix-until-green: implement, run the checks, fix, run them again. Then a review, and a gate where you decide.
- Iteration 1Checks failed · fix
- Iteration 2Checks passed
One of 11 built-in flows. Flows are experimental: turn them on in Settings → Experimental.
Flows that loop until green

A fix-until-green run: the loop between checks and fix, then review and the gate. Review
Read the diff. Rewind one turn.
Every change is a diff before it's a commit. lisbon's fix is right, but its last turn also renamed a helper nobody asked about. You restore the checkpoint before that turn: the files come back, the chat picks up from there, and you ask again, smaller.
- Turn 3 Fix the redirect after sign-in
- Turn 4 Keep the session on refresh Restore to here
- Turn 5 Rename safeReturnTo everywhere
Review every change

The Editor layout: lisbon's diff in the middle, its chat beside it. Ship
Open the pull request from your terminal.
Your hands are already on the keyboard. Almost everything in the window also works from the
foremancommand, and every command takes--jsonwith stable exit codes, so the same line works in a script or in CI.foreman pr create atlas --draftA real command line
~/code/atlas fix-login-redirect $ foreman pr create atlas --draft ✓ Draft pull request opened: acme/atlas#142 fix-login-redirect → main · 3 files · +7 −3 $ foreman approvals --json [] $An illustration of opening a draft pull request with the foreman command. Dusk
Keep the branch fresh while you sleep.
kyoto's launch branch has to live until Thursday. Before you go, you put the built-in keep-branch-fresh flow on a schedule. Overnight it merges origin/main, fixes what fails, and waits for you to look before it pushes.
Schedules run while Foreman is running and the computer is awake.
foreman service installstarts it with no window when you log in.Automations and schedules

Automations: templates to run on demand, and schedules beside them.
Done for the day. Every push today went through you.



Needs you 1
osaka Codex · Push
Push add-rate-limit to origin
Leaves this computer
Pushes 2 commits to origin (github.com/acme/atlas) as @you. It can't be undone.
Approved. osaka pushed 2 commits.
Held. The commits stay on this computer.
Paired · Operate · same sign-in as your desk



~/code/atlas fix-login-redirect
$ foreman pr create atlas --draft
✓ Draft pull request opened: acme/atlas#142
fix-login-redirect → main · 3 files · +7 −3
$ foreman approvals --json
[]
$ 
The short version
What does an AI pair programming workflow look like?
An AI pair programming workflow is a loop: you give a coding agent a scoped task on its own branch, answer its questions, approve anything risky, review the diff, and ship a pull request. Foreman runs that loop for several agents at once, each in its own git worktree, from one window on your computer.
- Scope. One task per workspace, one worktree and branch per agent.
- Watch. Tiles shows every agent; step in only where you're needed.
- Approve. Anything that leaves your computer waits for you, spelled out.
- Away. Answer approvals from your phone.
- Loop. A flow fixes and re-runs the checks until they pass.
- Review. Read the diff; restore a checkpoint when a turn goes too far.
- Ship. Open the pull request from the window or the terminal.
- Repeat. Schedules keep work moving while you're away.
Why does every agent get its own git worktree?
A git worktree is a second checkout of the same repository on a different branch. Giving each agent its own worktree means agents never overwrite each other's files, every change lands on a reviewable branch, and your main checkout stays untouched.
Can I approve an agent's actions from my phone?
Yes. Foreman's phone view shows what needs you and updates live, so you can answer approvals away from your desk. You pair the phone once with a six-digit code and choose whether it may watch or operate.
Does my code leave my computer?
Only through the agents you choose. Foreman runs locally with no account and no telemetry upload; your agents talk to their own model providers. In your own workspaces, an approval marks any command that leaves your computer before it runs, and AI employees hold Slack posts and Linear changes for you.
Side by side
How is Foreman different from Conductor and T3 Code?
Foreman is open source under Apache-2.0 and runs on Windows as well as macOS, while Conductor is a closed-source Mac app. Next to both, it adds flows that loop until checks pass, AI employees and a command line where every command takes --json.
—A dash means not listed on their site
| Feature | Foreman | Conductor | T3 Code |
|---|---|---|---|
| Open source | Apache-2.0 | No | MIT |
| macOS | Yes | Yes | Yes |
| Windows | Yes | Mac only | Yes |
| One git worktree per agent | Yes | Yes | Yes |
| Agents | Claude Code, Codex, Cursor, OpenCode, any ACP agent | Claude Code, Codex, Cursor | Claude Code, Codex and more |
| Flows that loop until checks pass, with human gates | Yes | — | — |
| AI employees with git-backed memory | Yes | — | — |
| Scheduled automations and auto-fix CI | Yes | — | — |
| Your own cloud sandboxes (E2B, Daytona, Vercel, Docker) | Experimental | — | — |
| Scriptable CLI, --json on every command | Yes | — | — |
| Approve from your phone | Yes | — | Yes |
| No account, no telemetry upload | Yes | — | — |
Tomorrow, 08:59
Get started in a minute
Download Foreman
The .dmg for Apple Silicon or the Windows installer. No Node.js, no checkout, no account.
Add a repo, sign in an agent
Add a local git repository, then sign in Claude Code, Codex, Cursor or OpenCode in Settings → Agents with your own login or API key.
Create workspaces
One per task. Each is a git worktree on its own branch. Send a prompt, open Tiles, and pour a coffee.
Prefer a terminal? Install the foreman command from Settings → Advanced, then:
- Start a workspace from a template and wait for the turn
foreman run fix-a-bug "the login page loops after sign-in" --repo atlas --wait - Follow up in the same workspace
foreman send atlas "also add a test for an expired token" - See what is waiting for you
foreman approvals - Open a draft pull request
foreman pr create atlas --draft - Loop until the checks pass
foreman flow run fix-until-green --input task="make the tests pass" --wait
After hours
Questions developers ask
What is Foreman?
Foreman is a free, open-source desktop app for macOS and Windows that runs AI coding agents in parallel. Claude Code, Codex, Cursor and OpenCode each work in their own git worktree on their own branch, while you review diffs, answer questions and open pull requests from one window.
Is Foreman free and open source?
Yes. Foreman is free and licensed under Apache-2.0, with the full source on GitHub at github.com/elite-build/foreman. There is no paid tier, no account and no Foreman cloud service.
Which coding agents does Foreman support?
Foreman runs Claude Code, Codex, Cursor and OpenCode, plus any agent that speaks the Agent Client Protocol (ACP). Each agent signs in with your own subscription or API key, set in Settings → Agents.
How do I run multiple Claude Code agents in parallel?
Add a git repository to Foreman and create one workspace per task. Each workspace is a separate git worktree on its own branch, so several Claude Code sessions edit the same repository at once without conflicts. Tiles shows them side by side.
Why does each agent need its own git worktree?
A git worktree is a second checkout of the same repository on a different branch. Giving each agent its own worktree means agents never overwrite each other's files, every change lands on a reviewable branch, and your main checkout stays untouched.
Does my code leave my computer?
Only through the agents you choose. Foreman runs locally with no account and no telemetry upload; your agents talk to their own model providers. In your own workspaces, an approval marks any command that leaves your computer before it runs, and AI employees hold Slack posts and Linear changes for you.
Do coding agents need my approval for every command?
Only if you want them to. Approvals are a setting: with Claude Code tool approvals on (Settings → Account), each command is spelled out part by part before it runs, anything that leaves your computer is marked, and Always allow shows exactly what it adds. Every agent still works in its own worktree, and checkpoints rewind any turn.
Which operating systems does Foreman run on?
Foreman runs on macOS (Apple Silicon, signed and notarized) and Windows. You need Git on your PATH and a login or API key for at least one agent.
Can Foreman run coding agents on a schedule?
Yes. Any saved prompt template can run hourly, daily, weekly or monthly in your time zone, and opt-in auto-fix sends failing CI logs to your agent. Scheduled runs use Claude Code and happen while Foreman is running, which can be without a window or from login.
Can I control Foreman from my phone?
Yes. Pair a phone with a QR code to answer approvals, steer agents, comment on diffs and read terminals, live. On the same Wi-Fi it connects directly, and anywhere else through your own Tailscale network. Native iOS and Android apps are coming soon.
Can coding agents run in a cloud sandbox like E2B?
Yes, as an experiment. A workspace's worktree and agent can run in a sandbox on your own E2B, Daytona, Vercel Sandbox or Docker account, for Claude Code, Codex and OpenCode. Provider keys stay on your computer and never go into the sandbox.
Which integrations does Foreman have?
GitHub (pull requests, merges, CI and reviews), GitLab, Linear, Slack and Langfuse, plus MCP. Each connection has a default per service, each repository can pick its own, and you choose what agents may do with it: nothing, read, or read and write.
How is Foreman different from Conductor?
Foreman is open source under Apache-2.0 and runs on Windows as well as macOS, while Conductor is a closed-source Mac app. Foreman also drives OpenCode and any ACP agent, and adds flows that loop until checks pass, AI employees and a scriptable command line.
How is Foreman different from T3 Code?
Both are open-source GUIs for coding agents. Foreman adds flows (agent graphs that loop until tests pass, with human gates), AI employees with a git-backed memory, a foreman CLI where every command takes --json, phone approvals and remote sandboxes.
Can I script Foreman or use it in CI?
Yes. The foreman command starts workspaces, sends messages, answers approvals, opens and merges pull requests and runs flows. Every command takes --json and exit codes are the same for every command.
Help build tomorrow's Foreman.
Foreman is three weeks old and built in the open: 31 releases since 1.0.0 on 27 October, Apache-2.0, no paid tier. Star it so other developers find it, fork it, or pick up a good first issue tonight.