Keep Visual Studio.
Give Copilot your machines.

AI Commander is not an alternative to Visual Studio, 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, JetBrains Junie, Kiro, Warp, Factory, Crush, LM Studio, Trae, Claude Desktop, n8n, OpenAI Agents SDK, JetBrains AI Assistant, or Open WebUI. 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.

Positioning in one line: Visual Studio and GitHub Copilot agent mode stay your IDE and agent; AI Commander is the hands on the box — build servers, GPU rigs, NAS, laptops — with no inbound ports and no exposed SSH.

Sibling harness pages: Claude Code · Codex · ChatGPT · Cursor · OpenCode · Factory · Kiro · Warp · Crush · LM Studio · Trae · Claude Desktop · n8n · OpenAI Agents SDK · JetBrains AI Assistant · Open WebUI · Gemini — and the rest on Connect your AI client.

Connect Visual Studio (GitHub Copilot agent mode)

In Visual Studio (2026, or 2022 version 17.14 and later), GitHub Copilot agent mode is the MCP client (Microsoft Learn: Use MCP servers in Visual Studio). It connects to remote HTTP servers by URL and to local stdio servers by command. The quickest path: in the Copilot Chat pane switch to Agent mode, open Tools, click + (in Visual Studio 2026, then Add custom MCP server), and enter a name plus AI Commander’s hosted endpoint https://aicommander.dev/mcp. Or put it in a .mcp.json file:

%USERPROFILE%\.mcp.json — remote MCP (HTTP)
{
  "servers": {
    "aicommander": {
      "type": "http",
      "url": "https://aicommander.dev/mcp"
    }
  }
}

%USERPROFILE%\.mcp.json loads the server for every solution; <solution>\.mcp.json scopes it to one repo and is usually checked in — so never put an API key or Authorization header in that one. Visual Studio also picks up .vscode\mcp.json and .cursor\mcp.json, so an existing AI Commander entry there may already show up.

Sign-in: Visual Studio supports OAuth for remote MCP servers per the MCP authorization spec. When the CodeLens above the server in .mcp.json shows Authentication Required (or Manage Authentication), select it and finish the one-time browser sign-in — AI Commander supports Dynamic Client Registration, so there is nothing to paste, and the connector is bound to your account so saved-machine aliases resolve. For a code-only first hour without signing in, use https://aicommander.dev/mcp?anonymous=1 as the URL instead.

Enable the tools, keep approvals on

Visual Studio disables a new server’s tools by default: open the Tools picker in agent mode and tick the AI Commander tools you want (for example list_machines, remote_exec, remote_job_start, remote_job_status, remote_job_logs). Copilot then asks before each tool call; the Confirm / Allow dropdown can remember a choice for the session, the solution, or always — keep remote_exec on per-call confirmation until you trust the workflow. Visual Studio 2026 (18.7+) also asks you to re-trust a server whose configuration or tool list changed. If your organization enforces a GitHub MCP allow list, an administrator has to allow the server first.

Prefer a local stdio bridge?

.mcp.json — stdio bridge
{
  "servers": {
    "aicommander": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "@aicommander/mcp"
      ],
      "env": {
        "AICOMMANDER_SERVER": "https://aicommander.dev"
      }
    }
  }
}

The bridge needs Node.js (npx) on your PATH; add "AICOMMANDER_TOKEN": "<your-key>" to env in your user-level file for alias-based access. Saving .mcp.json makes Copilot reload its servers. Long builds and test suites on a remote box belong in detached jobs (remote_job_start) so the chat does not have to wait.

Using Copilot in VS Code or the Copilot CLI instead? See Keep GitHub Copilot. Full client matrix (Visual Studio, JetBrains AI Assistant, Open WebUI, Claude Desktop, Claude Code, Codex, ChatGPT, Cursor, Copilot, Gemini, and more, plus REST): Connect your AI client. Menus differ between Visual Studio versions — check the linked Microsoft Learn article for yours.

What you get that the chat alone cannot

Not a replacement — a multiplier

You already useAI Commander adds
Visual Studio.mcp.json → servers.aicommander {"type": "http", "url": "https://aicommander.dev/mcp"}
JetBrains AI AssistantSettings → Tools → AI Assistant → MCP → Add → npx -y @aicommander/mcp (or a remote url)
Open WebUIAdmin → Integrations → External Tool Servers → MCP (Streamable HTTP), https://aicommander.dev/mcp, OAuth 2.1
Claude Codeclaude mcp add --transport http … → remote exec + jobs
Codex / ChatGPTSame MCP endpoint or REST /api/v1/exec
Cursor / Continue / Windsurfnpx @aicommander/mcp stdio bridge

FAQ

Is AI Commander an alternative to Visual Studio?
No. AI Commander is NOT an alternative to Visual Studio. It is a remote-machine harness. You keep Visual Studio and GitHub Copilot (or 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, JetBrains Junie, Kiro, Warp, Factory, Crush, LM Studio, Trae, Claude Desktop, n8n, OpenAI Agents SDK, JetBrains AI Assistant, or Open WebUI) as your AI agent. AI Commander gives those tools secure remote shell, detached jobs, and file transfer on machines you own — over MCP or REST. It is also not an alternative to Claude Code, Codex, or ChatGPT; it works with them the same way.
How do I connect Visual Studio to AI Commander?
In Copilot Chat switch to Agent mode, open Tools, click + and add a server with URL https://aicommander.dev/mcp. Or add servers.aicommander with type http and that url to %USERPROFILE%\.mcp.json (all solutions) or <solution>\.mcp.json. Select the Authentication Required CodeLens to finish the one-time browser sign-in, then enable the AI Commander tools in the Tools picker. Use ?anonymous=1 for a code-only first hour. See /howto/ for the full matrix.
Is this the same as the GitHub Copilot page?
No. /with-copilot/ covers the Copilot CLI and VS Code. This page covers the Visual Studio IDE on Windows: Copilot agent mode, .mcp.json, CodeLens sign-in, and the Tools picker. Both reach the same AI Commander machines.
Why are the AI Commander tools greyed out in Visual Studio?
Visual Studio adds a new MCP server with its tools disabled. Open the Tools picker in agent mode and enable the ones you need. If the server itself is blocked, your organization may enforce a GitHub MCP allow list that an administrator has to update.

Next steps

Install the agent on a machine you control, add https://aicommander.dev/mcp to Visual Studio’s .mcp.json, enable the tools in agent mode, then ask Copilot for a real task on a remote box.