Keep Warp.
Give it your machines.
AI Commander is not an alternative to Warp, Claude Code, Codex, ChatGPT, Cursor, GitHub Copilot, Sourcegraph Cody, Grok, Gemini, Windsurf, Continue, Cline, Roo Code, Zed, OpenCode, Qwen Code, goose, Amp, Kilo Code, OpenHands, Amazon Q Developer, Augment Code, Devin, or JetBrains Junie. It is the remote-machine harness those tools plug into over MCP: outbound-only shell, detached jobs that outlive the chat, and file transfer on hardware you already own.
- Install the host agent on the machine you want your AI to control. It shows a private machine code.
- Connect your AI: paste this prompt, then replace
AIC-…with your code —Help me connect this AI client to AI Commander using https://aicommander.dev/start.md. Then check the status of my machine: AIC-…— or add the connector URLhttps://aicommander.dev/mcpand sign in with an email code (per-client steps).
Sibling harness pages: Junie · Cursor · GitHub Copilot · Cody · OpenHands · Amazon Q · Augment · Devin · Aider · Cline · Roo Code · Zed · OpenCode · Qwen Code · goose · Amp · Kilo Code · Continue · Windsurf · Claude Code · Grok · Codex · ChatGPT · Gemini · Cua Spaces · Kiro.
Connect Warp (remote MCP)
Warp, the agentic development environment built around the terminal
(Warp MCP docs),
lets its local agents use MCP servers added under Settings → Agents → MCP servers (or Warp Drive → MCP Servers).
Per Warp’s docs, a Streamable HTTP or SSE Server (URL) needs only a url (plus optional headers),
and Warp opens a browser sign-in for OAuth-protected servers and stores the credentials on your device.
Click + Add and paste:
{
"aicommander": {
"url": "https://aicommander.dev/mcp"
}
}
Prefer a file that follows you across repos? Warp also reads file-based servers from ~/.warp/.mcp.json (global, auto-spawned by default)
or .warp/.mcp.json at the project root (project scope). The built-in
/agent-add-mcp skill can write that entry for you, with your approval.
Do not commit secrets: if you add an Authorization header with an API key, keep it in the global file or in Warp settings.
Prefer a local stdio bridge?
{
"mcpServers": {
"aicommander": {
"command": "npx",
"args": ["-y", "@aicommander/mcp"],
"env": { "AICOMMANDER_SERVER": "https://aicommander.dev" }
}
}
}
Running Warp’s cloud agents instead of local ones? They take MCP servers from the CLI or an agent config file, and per Warp’s docs a raw
OAuth url without an Authorization header cannot be completed mid-run — pass a Bearer API key in headers
or set the server up as a managed installation first (see Warp’s cloud-agent MCP docs).
Full client matrix (Warp, Kiro, Junie, Augment, Devin, Claude Code, Codex, ChatGPT, Cursor, Copilot, Cody, Grok, Gemini, OpenCode, Qwen Code, goose, Amp, Kilo Code, Continue, Windsurf, Cline, Roo Code, Zed, OpenHands, Amazon Q, REST): Connect your AI client.
Over the remote URL, first connect may open a one-time browser sign-in (OAuth Dynamic Client Registration).
The npx bridge does not sign in through a browser: without AICOMMANDER_TOKEN in its env it connects anonymously with a code-only first hour; set it to your API key to bind it to your account.
On the remote URL, append ?anonymous=1 to skip the sign-in, for a code-only first hour.
For optional account features (saved aliases), pass a Bearer API key in headers.
Menu names and options can differ between Warp versions — check the linked Warp docs for your build.
What you get that the chat alone cannot
- Remote shell on Linux, macOS, and Windows — by session code or saved alias.
- Detached jobs for builds, test suites, and training on the box that has the repo.
- Fleet list + GPU facts so Warp picks the right machine before it burns an hour.
- File courier (Pro) for checkpoints and diffs up to 100 MiB.
Not a replacement — a multiplier
| You already use | AI Commander adds |
|---|---|
| Warp | Settings → Agents → MCP servers → + Add → https://aicommander.dev/mcp (or ~/.warp/.mcp.json) |
| Kiro | mcpServers.aicommander.url in ~/.kiro/settings/mcp.json |
| JetBrains Junie | mcpServers.aicommander.url in ~/.junie/mcp/mcp.json |
| Augment Code | auggie mcp add aicommander --transport http --url https://aicommander.dev/mcp |
| Devin | Settings → MCP → custom HTTP → https://aicommander.dev/mcp |
| Claude Code | claude mcp add --transport http … → remote exec + jobs |
| Codex / ChatGPT | Same MCP endpoint or REST /api/v1/exec |
| Cursor / Continue / Windsurf | npx @aicommander/mcp stdio bridge |
How Warp handles MCP servers
- Servers are managed under Settings → Agents → MCP servers, or in Warp Drive → Personal → MCP Servers. Each one is either a CLI server (a command plus optional
working_directory) or a URL server (Streamable HTTP or SSE, withurlandheaders). - Warp also reads file-based configs:
~/.warp/.mcp.jsonglobally and.warp/.mcp.jsonper project. It can pick up servers you already set up for Claude Code, Codex or in~/.agents. - Global servers in
~/.warp/.mcp.jsonstart on their own. Global servers picked up from Claude Code, Codex or other agents only start automatically after you turn on "Auto-spawn servers from third-party agents". Project-scoped servers have to be started by hand in each session. - Servers you added in settings that were running when you quit Warp start again on the next launch.
- The
/agent-add-mcpskill lets the agent add a server for you. - OAuth credentials are stored per device. On a new computer you sign in again, and
rm -rf ~/.mcp-authclears stuck tokens. Each server has a View Logs button for debugging.
Checked against Warp docs: MCP in November 2026. Client behaviour changes between releases; the official docs win if they disagree.
A workflow that suits Warp
Warp is a terminal, so AI Commander fits as the "other machines" half: the local shell stays in Warp's panes, and remote boxes go through MCP without SSH keys on the laptop.
- If AI Commander is already in your Claude Code or Codex config, let Warp detect it instead of adding it twice.
- Share the server with your team from Warp Drive. Share it without the
Authorizationheader and let each person add their own key. - Ask the agent: "Compare the running config on the three web nodes and show me what differs." It calls
list_machines, thenremote_execon each node.
Limits and gotchas in Warp
- A project-level
.warp/.mcp.jsonserver doesn't start by itself in a new session. Start it, or move AI Commander to the global file. - After switching computers, AI Commander asks you to sign in again, because Warp keeps OAuth tokens per device.
- Warp's docs say some models handle MCP better than others. If tools aren't called, try another model first.
FAQ
aicommander with url set to https://aicommander.dev/mcp. Or put the same entry under mcpServers in ~/.warp/.mcp.json. Prefer stdio? Use command npx with args -y @aicommander/mcp. The remote URL may ask for a one-time browser sign-in (OAuth); the npx bridge never does: it connects anonymously unless AICOMMANDER_TOKEN is set in its env. On the remote URL, use ?anonymous=1 for a code-only first hour. See /howto/ for the full matrix..warp/.mcp.json. Project-scoped servers need a manual start each session; servers in ~/.warp/.mcp.json start automatically.Next steps
Install the agent on a machine you control, add AI Commander in Warp’s MCP servers page (remote URL or npx -y @aicommander/mcp), then ask Warp’s agent for a real task on a remote box.