Mobile Use
This server does not currently have a deployable release on Glama. Follow the repository's setup instructions to install and configure it for your MCP client, including any required credentials and local dependencies.
Mobile Use
让支持 MCP 的助手操作一台已授权的 Android 真机或模拟器。比如「打开计算器,算出 37 × 12」:助手先从控件树找到按键;只有界面没有提供足够的文字或控件时,才需要看截图。对于已经走过的路径,可以把几步导航交给本机一次执行。
接上第一台手机
准备好 uv 和 Android Platform Tools。手机开启 USB 调试并接受电脑的授权;无线设备先用 ADB 配对、连接。终端运行 adb devices -l,确认目标设备状态为 device。
克隆仓库并安装锁定的依赖:
git clone https://github.com/NingCold/mobile-use.git
cd mobile-use
uv sync --frozen使用 Codex:在刚克隆的目录下打开 PowerShell,注册本地服务。命令保存绝对路径,因此以后从其他项目打开聊天也能使用同一份安装。
$project = (Resolve-Path .).Path
codex mcp add mobile-use --env PYTHONIOENCODING=utf-8 -- uv run --directory "$project" --frozen mobile-use让宿主重新加载 MCP 或打开新聊天,调用 get_status 查看运行中服务的能力。其他支持 stdio 的宿主也可以使用同一条 uv run --directory <项目绝对路径> --frozen mobile-use 启动命令。可选的 Mobile Use Skill 说明助手该如何选择观察和动作;它不能代替服务注册。
先给助手一个只读任务:
用 Mobile Use 列出在线设备。连接我的手机,告诉我当前应用和页面上的主要控件,不要点击;然后断开连接。
如果列出了多台设备,指定本次要用的序列号。设备需保持解锁。一次只有一个 Mobile Use 会话可以控制同一序列号;结束时调用 disconnect 释放占用,不会关闭应用或断开无线 ADB。
Related MCP server: ARTEMIS MCP Server
怎么操作既快又稳
有文字的控件:用
observe(detail="brief", image="auto")看当前页,再把最新的observation_id和已看到的element_id交给tap。明确身份的按钮可在位置变化后重新定位;不会因为旁边的倒计时变了就把整页判为失效。自绘页面:先看图。已看清文字和所属卡片时,可用本地 OCR 定位;遇到多个匹配、遮挡或窗口变化时停止。不要猜坐标。
已知路径:用 navigate 声明允许的应用、页面条件和目标。刷新会移动按钮时,只等目标位置稳定。不要在每次切页后机械地追加一次观察。
异常或中断:先看动作是否已经发送。已发送或结果未知时,核对当前状态,不要重放整段。需要更多文字时,先用
get_observation展开最近一次的缓存;它不刷新屏幕,也不会延长观察有效期。
数据留在哪里
控件解析、OCR 和手势执行在本机完成;工具返回的文字和图片会进入所用 MCP 宿主。默认不把观察截图写入磁盘。仓库安装的审计记录默认在 artifacts/,可能包含界面文字、设备标识和操作轨迹;调试脚本还可能主动保存图片。不要把这些文件直接贴到公开 issue。服务可用 --artifacts 指定私有记录目录。
This server cannot be deployed
Maintenance
Related MCP Connectors
Drive real devices from your AI Coding tool. Embed a client SDK (Unity, Godot, Flutter, iOS/macOS, Android, React Native, Web) in your app, then capture screenshots, traverse the UI tree, inject taps and key events, and run automated test tasks on the physical device over a secure relay.
Launch and verify high-fidelity mobile simulations from an AI agent.
Let AI agents control Android and iOS devices on Mobilerun Cloud: screen, input, apps, workflows.
Give your AI agent a real cloud Android phone: push APKs, drive it, get screenshots and logs.
Related MCP Servers
- FlicenseNot gradedqualityBmaintenanceEnables an agent to control a real Android phone by reading structured UI elements and performing actions such as tapping, swiping, typing, and extracting fields from known app screens.1-
- AlicenseNot gradedqualityBmaintenanceEnables AI assistants and test suites to drive real Android devices or emulators through MCP, executing natural-language automation and testing workflows while collecting screenshots, Logcat output, and device state.Apache 2.0
- AlicenseBqualityCmaintenanceEnables any MCP assistant to see and operate an Android phone or emulator over plain adb, reading the accessibility tree to tap elements by name and also supporting screenshots, typing, swiping, app launching and installation, and logcat — with nothing installed on the device and no root required.13MIT
- AlicenseNot gradedqualityAmaintenanceEnables MCP-compatible AI assistants to see and control Android devices through screenshots, taps, swipes, typing, app management, UI inspection, file transfer, shell commands, and audio/video streaming. It connects over ADB with optional low-latency acceleration for faster input and near-instant screenshots.365 npmMIT