跳到正文
cccskills
免费在 GitHub 公开

cua-sandboxes

Create disposable Linux or macOS computers for running code, testing apps, operating desktop interfaces, and browsing websites. Use isolated environments locally or in the Cua cloud, then delete them when the work is complete.

中文简介正在准备中。当前显示英语简介。

原文查看原文说明

Create, use and clean up cua sandboxes (disposable Linux or macOS computers) locally or in the Cua cloud with the `cua` CLI or the cua SDK, and browse the web inside one. Use when a task needs an isolated machine to run code, test an app, drive a desktop GUI, browse or test a website, fill a web form, take a web page screenshot, or reproduce something without touching the user's own computer.

查看安装方法

适用场景

  • Run code in an isolated environment
  • Test an app or reproduce an issue
  • Fill web forms and capture page screenshots
  • Create and clean up temporary test environments

包含的文件(1)

  • SKILL.md5.7 KB

SKILL.md(原文)

安装前可以查看交给 AI 代理的指令内容。

cua sandboxes

A sandbox is a disposable computer: a container or VM on this machine or in the Cua cloud. The cua CLI and the cua SDK (Python, TypeScript, Swift, Kotlin, Rust) share one runtime and one list of sandboxes.

Before you start

cua --version            # the Rust cua CLI
cua auth status          # the cloud needs `cua auth login` (or CUA_CLIENT_ID/CUA_CLIENT_SECRET)
cua config get default.on   # where sandboxes run without --on (local unless changed)
cua runtime doctor       # local runtimes (containers, gVisor, QEMU, Lume); read-only

Local sandboxes need no account. Cloud sandboxes are metered: always delete them when done.

Create

cua sb create linux --name dev                 # local (default): a gVisor container
cua sb create linux --kind vm --name dev       # local Linux VM (QEMU)
cua sb create macos --name mac                 # local macOS VM (Apple silicon, Lume)
cua sb create linux --on cloud --name dev      # the Cua cloud
cua sb create ghcr.io/org/image:tag --name x   # any OCI image, local or --on cloud
cua sb create --on direct:127.0.0.1:3211 --token TOKEN --name mine   # an existing machine running cua-spacesd

--on is where (local, cloud, direct:<addr>), --kind what (auto, container, vm), --runtime which engine (auto; local gvisor, runc, qemu, lume; cloud gvisor, kubevirt). An impossible combination fails with invalid placement and lists the valid values. Local runtimes are set up on first use (cua runtime setup does it explicitly). Omit --name for a generated name. cua sb ls prints each sandbox's ref (local:dev, cloud:dev, direct:host:port); every NAME argument takes a ref or a bare name that is unique across locations (--local / --cloud narrow it). Add --json to any command for machine-readable output. exec, shell, cp, screenshots and GUI control go through cua-spacesd, so use an image that ships it (the linux alias does). Plain images still get lifecycle, logs and port-forward.

Use

cua sb ls                          # every location (--local or --cloud filters)
cua sb exec dev -- uname -a        # run a command
cua sb shell dev                   # interactive shell
cua sb screenshot dev              # save a screenshot
cua sb vnc dev                     # open the desktop in a browser

For GUI work, select the sandbox once and use cua do (see the gui-automation skill):

cua do switch dev
cua do screenshot
cua do click 400 300

Clean up

cua sb rm dev --force              # delete
cua fleet pools gc                 # remove idle auto-managed cloud capacity
cua sb suspend dev                 # or pause it and `cua sb resume dev` later

From code (Python)

import asyncio, cua

async def main():
    c = cua.embedded()
    sb = await c.sandboxes().create(cua.SandboxCreateOptions(
        on="local",
        image="ghcr.io/trycua/linux:24.04", name="dev"))
    env = await sb.spacesd(None)
    out = await env.run(cua.SpacesdCommand(program="uname", args=["-a"]))
    print(out.stdout.decode())
    await sb.delete()

asyncio.run(main())

The TypeScript, Swift and Kotlin bindings expose the same objects (Cua, sandboxes(), Sandbox.spacesd()).

From the cua MCP server

When the cua MCP server is connected, prefer its tools: images_list (which images exist), sandbox_list, sandbox_create, sandbox_get, sandbox_delete, then the computer_* tools (screenshot, click, type, shell, file) against the sandbox.

Browse the web

Browse inside a sandbox, never in the user's own browser. Everything below is cua MCP tool calls; no other browser tool or setup is needed.

  1. Pick an image: images_list {"browser": true} (the canonical ghcr.io/trycua/linux:24.04 ships Chromium).
  2. Create it with a browser: sandbox_create {"browser": true, "url": "https://example.com"} (add "on": "cloud" for the cloud). It returns id, session and browser.target_id / browser.tab_id.
  3. Drive the browser with call_tool {"space": <id>, "tool": <tool>, "arguments": {"session", "target_id", "tab_id", ...}}:
    • read: get_browser_state {"snapshot_format": "semantic_v2"} returns an outline and refs (p1:0); add "include_screenshot": true for an image;
    • browser_navigate {"url"}, browser_click {"ref", "delivery_mode": "foreground"}, browser_type {"ref", "text", "replace": true} (text "\n" with "mode": "keystrokes" presses Enter), browser_pointer (hover, scroll, drag);
    • read again after every page change: refs are per snapshot. To test a local app, run it in the sandbox (space_write, space_bash) and open http://127.0.0.1:<port>.
  4. Logged-in sites: only when the user asks. teleport_browser_session with the exact sites first returns what would move and a consent requirement; show it to the user and continue only after they approve. Never move a session without that approval.
  5. Clean up: sandbox_delete {"name": <id>}.

From a terminal: cua images ls --browser, cua sb create --browser --open https://example.com, cua sb view <id> to watch.

Rules

  • Never run destructive commands on the user's own machine when a sandbox would do.
  • Delete cloud sandboxes you created as soon as the task ends.
  • Report the sandbox name you used so the user can inspect or remove it.

评价

还没有评价。欢迎分享你的使用感受。

评论

无需登录即可发表。欢迎分享使用感受、提问,或说明在哪些环境下无法运行。

还没有评论。来写第一条吧。

提交后会立即公开。请勿填写邮箱、电话等个人信息。URL 最多 2 个。

0 / 1000 字

此表单受 reCAPTCHA 保护,适用 Google 的隐私权政策和服务条款。

同一仓库的技能

Operate native apps on macOS, Windows, and Linux using identified controls, menu paths, or screen coordinates. Target a specific window, verify the result after each action, and consult available recent activity when resuming work.

中文简介正在准备中。当前显示英语简介。

  • Operate a specific app window
  • Click controls and enter text
  • Adjust window positions and sizes
trycua/cua3万2026年10月18日 更新

Work in a local or remote computer environment whose screen you can watch. Run commands, transfer files, launch coding agents, move signed-in app sessions, and share the host computer’s network when needed.

中文简介正在准备中。当前显示英语简介。

  • Run tasks in an isolated environment you can watch
  • Transfer files into and out of a Space
  • Launch and monitor coding agents
trycua/cua3万2026年10月18日 更新

Save agent memory and outputs in Cua Volume, shared storage that persists beyond a Space, the working environment. Hand files to the user, read shared references, and check synchronization, version history, and conflicting edits.

中文简介正在准备中。当前显示英语简介。

  • Keep outputs after a working environment ends
  • Save agent memory for later work
  • Deliver files to the user
trycua/cua3万2026年10月18日 更新

Operate apps and websites by inspecting screenshots, clicking controls, typing, and dragging. Use it to fill forms, check layouts, and test complete user flows, with recorded actions available for replay and review.

中文简介正在准备中。当前显示英语简介。

  • Fill forms and use file upload dialogs
  • Test buttons and complete user flows
  • Check layouts and capture screenshots
trycua/cua3万2026年10月18日 更新

jev-use

免费

Build a computer-control loop where Cua Driver observes and acts, while TypeSafe Jev chooses from predefined actions. Validate each choice, execute one action at a time, and verify the result before continuing.

中文简介正在准备中。当前显示英语简介。

  • Build a Jev integration for browser automation
  • Create action candidates for native desktop apps
  • Reject stale or unknown action choices
trycua/cua3万2026年10月18日 更新

Review open GitHub issues, proposals, and pull requests to identify useful maintainer work. Get ranked candidates with evidence, dependencies, and validation needs, then begin work only on an item you explicitly select.

中文简介正在准备中。当前显示英语简介。

  • Choose the next maintenance task
  • Prioritize a backlog for the available work window
  • Find pull requests that need review
trycua/cua3万2026年10月18日 更新

查看 trycua 的全部技能

报告此技能的问题