ai-phone — AI Tool by dongxinsuperman

by dongxinsuperman · AI Tool · ★ 68

Last updated: · Indexed by AgentSkillsHub · Auto-synced every 8h

🔒 Is ai-phone safe to install? View the security audit →

About ai-phone

ai-phone 面向中小型公司的三端真机 AI 自动化中台 —— iOS / Android / HarmonyOS 同级原生支持,自然语言驱动的纯视觉决策,开箱即用的调度队列与多设备并发,执行器可插拔,一台 Mac 即可起完整链路。新 Mac 从 0 到 1 的完整部署见 deployment-from-zero(从0到1部署指南)。 产品形态:ai-phone 不是一个执行器 SDK,而是把"投递批次 → 设备池调度 → 自然语言执行 → 终态广播 + HTML 报告 + 大盘统计"做成 QA 团队 / 业务回归大盘开箱即用的中台能力。执行器是其中一个可替换组件:默认内置自研的 VLM 纯视觉决策循环,并在外层包了页面稳定、卡死检测、审判、断言、状态路标、瞬态 UI 处理等执行安全层;也可挂载第三方执行器作为额外引擎选项。 产品边界与出发点:ai-phone 不是孤立的“点手机工具”,也不是测试平台本身,而是 AI 测试链路里的真机执行层。上游可以由 Cursor、CI、内部平台或 AI 助手生成可消费测试 case,下游由 ai-phone 把这些自然语言 case 投递到真实 iOS / Android / HarmonyOS 设备上执行,并回传截图、日志、报告和终态结果。详细说明见 product-boundaries(产品边界)。 看一眼实际样子 按"调度 → 调试 → 决策护城河 → 产出 → 观测"5 个节点串起来看: 1. 调度队列 —— 三端独立 FIFO + 正在执行 + 最近批次状态一栏拉通,外部投递的 Submission 实时分发到设备池: 2. 单设备调试 —— 浏览器即客户端:左实时画面(scrcpy / WDA / hypium 三端原生推流)、右自然语言 Goal 输入 + 步骤日志面板,回写通道与 VLM 共用,业务测试同学零安装、零配置: 3. 辅助系统护城河 —— ai-phone 不是把截图直接交给 VLM 赌下一步,而是在执行外层加了一套可监督的安全层:页面稳定后再决策,本地 pHash 捕捉同坐标 / 同屏 / 滑动震荡,异常自动召唤审判系统

ai-automationandroidharmonyosiosmobile-automationmobile-testingreal-devicescrcpyui-automationvisual-automation

Quick Facts

Stars68
Forks17
LanguagePython
CategoryAI Tool
LicenseMIT
Quality Score54.9713875372186/100
Open Issues8
Last Updated2026-08-07
Created2026-04-28
Platformsbrowser, python
Est. Tokens~16k

Compatible Skills

These tools work well together with ai-phone for enhanced workflows:

  • android-agent — semantic(0.28)+complementary+rare_topics+same_lang+similar_pop+shared_platform (59%)

ai-phone alternative? Top 6 similar tools

Looking for a ai-phone alternative? If you're comparing ai-phone with other ai tool tools, these 6 projects are the closest alternatives on Agent Skills Hub — ranked by topic overlap, star count, and community traction.

  • android-agent by ghost-in-the-droid · ⭐ 277

    Open-source framework to drive a real phone with AI agents: Android over ADB, iPhone over WebDriverAgent, 62 M

  • native-devtools-mcp by sh3ll3x3c · ⭐ 107

    MCP server for computer use & browser automation - screenshot, OCR, click, type, find_text, Chrome/Electron C

  • AppClaw by appclawhq · ⭐ 100

    AI-powered mobile automation agent — describe what you want in plain English, AppClaw reads the screen, reason

  • AppClaw by AppiumTestDistribution · ⭐ 91

    AI-powered mobile automation agent — describe what you want in plain English, AppClaw reads the screen, reason

  • Autofish by memohai · ⭐ 108

    An AI Agent tool to automate Android devices reliably

  • scrcpy-mcp by JuanCF · ⭐ 53

    MCP server for Android device control via ADB and scrcpy — gives AI agents vision and control over Android dev

More AI Tool Tools

Explore other popular ai tool tools:

View all AI Tool tools →

Popular Python Agent Tools

Frequently Asked Questions

What is ai-phone?

ai-phone is iOS / Android / HarmonyOS 三端真机 + 三端虚拟机 AI 自动化中台 · 自然语言驱动纯视觉执行,开箱即用的设备调度、多设备并发与报告回传. It is categorized as a AI Tool with 68 GitHub stars.

What programming language is ai-phone written in?

ai-phone is primarily written in Python. It covers topics such as ai-automation, android, harmonyos.

How do I install or use ai-phone?

You can find installation instructions and usage details in the ai-phone GitHub repository at github.com/dongxinsuperman/ai-phone. The project has 68 stars and 17 forks, indicating an active community.

What license does ai-phone use?

ai-phone is released under the MIT license, making it free to use and modify according to the license terms.

What are the best alternatives to ai-phone?

The top alternatives to ai-phone on Agent Skills Hub include android-agent, native-devtools-mcp, AppClaw. Each offers a different approach to the same problem space — compare them side-by-side by stars, quality score, and community activity.

View on GitHub → Browse AI Tool tools