Install mcp-qemu-lab

Pick your client, copy the command, done.

Add it to Claude Code

claude mcp add -e "MCP_QEMU_LAB_WORKSPACE=${MCP_QEMU_LAB_WORKSPACE}" qemu-mcp -- uvx --from git+https://github.com/Kevin4562/QEMU-MCP.git@main mcp-qemu-lab

Required:MCP_QEMU_LAB_WORKSPACE+ 1 optional

Environment Variables

Set these before running mcp-qemu-lab.

VariableDescriptionRequired

MCP_QEMU_LAB_WORKSPACEAbsolute path to the directory for storing VM workspaces, logs, and artifacts.Yes

MCP_QEMU_LAB_RUN_INTEGRATIONEnables integration tests (set to 1).No

Available Tools (7)

Once configured, mcp-qemu-lab gives your AI agent access to:

vm_createCreates a new QEMU virtual machine.

nameimage_path

vm_startStarts a specified QEMU virtual machine.

name

vm_stopStops a specified QEMU virtual machine.

name

guest_execExecutes a command inside the guest virtual machine.

vm_namecommand

guest_dump_memoryPerforms a full RAM dump of the guest virtual machine.

vm_nameoutput_path

process_dump_coreGenerates a core dump for a specific process inside the guest.

vm_namepid

debugger_attachAttaches a debugger to a process within the guest.

vm_namepid

Try It Out

After setup, try these prompts with your AI agent:

→Create a new QEMU virtual machine and start it.

→List all running processes inside the guest VM.

→Attach the debugger to the process with PID 1234 and read its registers.

→Perform a full memory dump of the current guest VM and save it to the workspace.

→Execute 'ls -la /etc' inside the guest and return the output.

Prerequisites & system requirements

An MCP-compatible client (Claude Code, Cursor, Windsurf, Claude Desktop, or Codex)
Python 3.8+ with pip installed
MCP_QEMU_LAB_WORKSPACE — Absolute path to the directory for storing VM workspaces, logs, and artifacts.

Keep this setup from going cold

Save the docs, env vars, and workflow around mcp-qemu-lab in Conare so Claude Code, Codex, and Cursor remember it next time.

Remember this setup

← mcp-qemu-lab overview Browse other servers →