LoopyLuci

VM Harness — DevOps skill for Claude Code

DevOps community

VM Harness allows agents control over QEMU, VMware, VirtualBox, Docker, Kubernetes and VMs and containers safely and securely with VM streaming between devices and the mobile app for Android able to p.

How to install VM Harness

This entry records only its repository, not the path inside it, so there is no exact command to give. Open LoopyLuci/VM-Harness and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What VM Harness does

VM Harness allows agents control over QEMU, VMware, VirtualBox, Docker, Kubernetes and VMs and containers safely and securely with VM streaming between devices and the mobile app for Android able to pair to a desktop install and manage it as well as use VMs streamed to the mobile device.

Alternatives in DevOps

  • Appwrite — Appwrite® - complete cloud infrastructure for your web, mobile and AI apps 57.5k ★
  • Codeg — Aggregate and browse AI coding agent sessions (Claude Code, Codex, Gemini CLI, etc.) in one place 850 ★
  • Cheetahclaws — CheetahClaws: A Fast and Easy-to-Use Agent Harness Infrastructure for Long-Horizon, Multi-Model, and Tool-Usin 768 ★

README

VM-Harness

Control virtual machines and containers from one place: a desktop window, a local API, an MCP server for AI agents, and a command line. QEMU, VirtualBox, VMware, Hyper-V, WSL and KVM; Docker, Podman, Kubernetes and Compose.

Quick start

pip install -e ".[dev]"
python -m vm_harness gui           # the window
python -m vm_harness mcp           # MCP server for Claude Desktop, ABP, Cursor... (stdio)
python -m vm_harness call vm.list  # any operation from the command line

Everything goes through one local service, the hub: see **[docs/control.md](docs/control.md)** for the API, the 160+ operations, MCP, and driving the window remotely.

Features

  • 34 GUI Panels — Complete VM control from a polished dark-themed desktop interface
  • CLI Tool — Command-line interface for scripting and agent access
  • REST API — HTTP endpoints for programmatic control
  • MCP Server — every operation as an MCP tool, including driving the window (python -m vm_harness mcp)
  • Self-Healing — Atomic state, process guardian, automatic recovery
  • Multi-VM — Control multiple QEMU instances from one interface
  • ISO Manager — Internal storage + external folder scanning + downloads
  • Guest Integration — SSH terminal, file browser, process manager

Quick Start

pip install -e ".[dev]"

qemu-mcp gui              # Launch GUI
qemu-mcp status           # CLI status
qemu-mcp api start        # REST API
python -m vm_mcp          # MCP server

GUI Panels (23 Total)

# Panel Description
1 Dashboard Live VM status — state, PID, resources, activity log, quick actions
2 VM Switcher Multi-VM management — add/remove/switch between VMs
3 VM Control Lifecycle — start/stop/reset/suspend/resume/eject ISO
4 Guest Terminal SSH command execution with history
5 Guest Agent Guest processes, services, files via SSH
6 Telemetry Real-time C