# 使用 Homebrew 安装 lume

查看 lume 的安装路径、可执行文件、元数据以及面向 AI 代理工作流的安全说明。

## 安装

```sh
sudo av install brew:lume
```

其他安装命令:

### macOS

- Homebrew (100%):

```sh
brew install lume
```

  证据: local Homebrew formula metadata

## 软件包事实

- **软件包键:** brew:lume
- **软件包管理器:** Homebrew
- **版本:** 0.5.1
- **来源摘要:** Create and manage Apple Silicon-native virtual machines
- **主页:** <https://cua.ai>
- **仓库:** <https://github.com/trycua/cua>
- **最后更新:** 2026-07-31T12:53:37Z
- **已生成:** 2026-08-03T19:37:03+00:00

## 可执行文件

- lume (别名)

## 安装行为

- Bottle: 不可用

## 版本和新鲜度

- 页面生成时间: 2026-08-03
- 管理器版本: 0.5.1
## 项目历史与用法

Lume is the Cua project's Apple-Silicon-focused VM runtime and CLI for creating and managing macOS and Linux virtual machines with Apple's Virtualization Framework.

### 项目历史

Lume lives inside the `trycua/cua` repository, which was created publicly in January 2025 as open-source infrastructure for computer-use agents, sandboxes, SDKs, and benchmarks. The Lume README describes it as a CLI and framework for macOS and Linux VMs using Apple's Virtualization Framework.

The official Lume introduction frames the project around agent workloads, CI/CD, and macOS automation. It exposes a single binary with an HTTP API, can create and run VMs headlessly, and is designed to be used directly from the CLI or programmatically through `lume serve`.

### 采用历史

Lume's adoption is tied to the rapid growth of computer-use agents and the practical need for reproducible macOS sandboxes on Apple Silicon. The Cua repository accumulated substantial GitHub attention during 2025-2026, and the README positions Lume alongside Cua drivers, sandbox SDKs, and benchmarks as one of the monorepo's core packages.

Homebrew packages the `lume` binary, while upstream docs also advertise install scripts and VM image pull/run workflows. The project is newer than traditional VM managers, so its package history is still forming around AI-agent, local CI, and macOS automation use cases.

### 使用方式

The official docs show `lume create test-vm --os macos --ipsw latest` followed by `lume run test-vm`, and the repository README shows pulling and starting a macOS VM image with `lume run macos-sequoia-vanilla:latest`.

Lume is intended for testing across macOS versions, automating macOS tasks, running local CI/CD, sandboxing risky operations, and building AI agents that interact with virtualized desktops through the broader Cua stack.

### 为什么软件包爱好者会关心

Lume is significant to package nerds because it packages a native macOS virtualization workflow into a CLI, not a heavyweight desktop app. Its constraints are also unusually concrete: Apple Silicon only, Apple's Virtualization Framework, IPSW-based macOS setup, sparse VM storage, headless operation, registries such as GHCR or GCS, and integration with agent tooling.

### 时间线

- 2025: Cua repository created publicly.
- 2025-2026: Lume documented as the monorepo's macOS/Linux VM management package.
- 2026: Cua repository shows hundreds of releases across monorepo components and active Lume documentation.
- 2026: Homebrew formula available for the `lume` CLI.

### Related projects

- Cua Computer SDK uses Lume's API for macOS automation workflows.
- Apple Virtualization Framework is the native virtualization layer Lume builds on.
- The Cua monorepo also includes Cua drivers, Cua sandbox packages, and Cua Bench.

### 来源

- <https://api.github.com/repos/trycua/cua>
- <https://api.github.com/repos/trycua/cua/releases>
- <https://cua.ai/docs/lume/guide/getting-started/introduction>
- <https://formulae.brew.sh/formula/lume>
- <https://github.com/trycua/cua>
- <https://github.com/trycua/cua/tree/main/libs/lume>


## 安全说明

formula declares a Homebrew service. infrastructure mutation or orchestration signal.

- **Geiger 风险:** orange / 中
- formula declares a Homebrew service
- infrastructure mutation or orchestration signal


## Combined YAML source

View the package source record on GitHub. [combined/lume.yml](https://github.com/mxcl/pkgdb/blob/main/combined/lume.yml)


## 来源

- pkg.so package database
- Geiger risk classifier
- curated package history
- pkgdb category and tag curation
- cross-ecosystem install command graph
