Keep n8n.
Give its AI Agent your machines.

AI Commander is not an alternative to n8n, 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, or OpenAI Agents SDK. 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.

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

Connect n8n (MCP Client Tool)

n8n’s MCP Client Tool node lets an n8n AI Agent call the tools of an external MCP server. In your workflow, open the AI Agent node, add a Tool, pick MCP Client Tool, and fill in:

Endpoint:          https://aicommander.dev/mcp
Server Transport:  HTTP Streamable
Authentication:    MCP OAuth2   (or Bearer Auth with an AI Commander API key)
Tools to Include:  Selected → list_machines, remote_exec, remote_job_start,
                   remote_job_status, remote_job_logs, session_status

MCP OAuth2: n8n’s MCP credential has Use Dynamic Client Registration on by default, and AI Commander supports it, so there is no Client ID to paste — you sign in once in the browser. AI Commander accepts HTTPS or loopback (http://localhost / http://127.0.0.1) redirect URLs, so the OAuth Redirect URL n8n shows must be HTTPS unless n8n runs on localhost — a self-hosted n8n served over plain HTTP on a LAN IP or hostname will not register. Bearer Auth: generate a personal API key in the AI Commander dashboard and store it in an n8n Bearer credential — never in a node parameter. Older MCP Client Tool versions only had an SSE endpoint; use a current n8n build, since AI Commander speaks Streamable HTTP.

Workflows run unattended — plan for that

By default an AI Commander API key or OAuth connector only acts while you have opened the dashboard within the last 24 hours, which limits the blast radius of a leaked credential. A scheduled n8n workflow at 3 a.m. will hit that gate. Either open the dashboard daily, or turn the gate off for your account under Credential security in the dashboard (this trades that protection away) — see Connect your AI client.

The node’s Timeout option defaults to 60 000 ms per tool call. For builds, backups, or training runs that take longer, have the agent start them with remote_job_start and check back with remote_job_status / remote_job_logs instead of waiting inside one remote_exec call. Use Tools to Include → Selected (or All Except) to expose only the tools a given workflow needs.

Want a fixed step without an LLM deciding? Recent n8n versions also ship a standalone MCP Client node (MCP Endpoint URL + Tool + inputs) that calls one tool directly — for example a nightly remote_exec health check on a saved alias.

Full client matrix (n8n, Claude Desktop, OpenAI Agents SDK, Claude Code, Codex, ChatGPT, Cursor, Copilot, Gemini, Trae, Crush, LM Studio, and more, plus REST): Connect your AI client. Prefer plain HTTP from an n8n HTTP Request node? The REST API is on the same page (REST API). Node options can differ between n8n versions — check the linked n8n 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 your n8n agent picks the right machine before it burns an hour.
• File courier (Pro) for checkpoints and diffs up to 100 MiB.
• Outbound-only agent on each machine — no inbound ports, no exposed SSH, no VPN to set up.

Not a replacement — a multiplier

FAQ