Essentials

Meta VR CLI overview

Updated: Aug 31, 2026

Overview

Meta VR CLI lets you drive your Meta Quest headset from a terminal. It also lets an AI coding agent drive it for you. The command is metavr.

It is one tool with two front ends. You can run commands yourself. Or your AI agent can call the same commands through a server that speaks the Model Context Protocol (MCP). Both reach the same code, so your choice of front end does not limit what you can do.

Note: Meta VR CLI was formerly named Horizon Debug Bridge (hzdb). The hzdb command still works as an alias. The @meta-quest/hzdb npm package still works too. Both are permanent, so older scripts keep running.

How it works

The logic lives in one place. The command line and the MCP server are two doors into it. That keeps the two in step as the tool grows.

Between them they let you:

Find headsets, check their state, and connect over Wi-Fi

Install, launch, and stop apps

Move files, read logs, and run shell commands

Take screenshots and drive the on-screen interface

Capture and analyze performance traces

Search the docs and the API reference

Search Meta’s 3D asset library

Publish builds to the Meta Horizon Store

Install other developer tools and simulators

Driving it yourself

Run metavr in a terminal and pick each command:

metavr device list

This suits scripts, automation, and any job where you want each step spelled out.

Letting your agent drive it

Your agent can run the command line directly. Describe what you want and it picks the commands, so you never name one yourself.

Meta VR CLI also runs a Model Context Protocol (MCP) server. For an agent that speaks MCP, this exposes the same capabilities as tools the agent discovers on its own, and the set stays current as the CLI updates.

Skills

Skills are reusable instructions. They teach an agent how to run a Quest workflow from end to end. Guided setup installs the Meta VR CLI skill next to the MCP server config, so your agent gets workflow context on top of the raw commands.

Key considerations

Consideration	Details
Signing in	Installing and using the CLI works without an account, and `metavr init` offers to sign you in. Signing in is worth doing: it unlocks Meta Horizon Store commands and other account-backed features, and more will depend on it over time. It opens a browser.
Installing	A native installer covers macOS, Linux, and Windows. You can also run the CLI through npm.
Updates	A binary from the native installer updates itself. One from a package manager defers to that package manager.

When to use Meta VR CLI

Use it whenever your work touches a headset. That covers installing builds, reading logs, capturing traces, taking screenshots, searching docs or assets, and shipping to the Meta Horizon Store.

Run it yourself when you want each step under your control. Let your agent run it when you prefer to describe the goal and have the agent pick the steps. Both reach the same commands.

Install Meta VR CLI: install the CLI and connect it to your AI coding agent.

Work with devices and apps: connect a headset, install builds, and read logs.

Profile and debug performance: capture and analyze a trace.

Set up your environment and tools: install developer tools and simulators, then configure settings and sign-in.

Meta VR CLI in other tools: the Meta tools that already bundle the CLI.

Troubleshoot Meta VR CLI: agent connection, sign-in, and update problems.

AI tooling overview: the wider set of AI tools for Meta VR development.