android-agent — security grade SAFE, quality 79/100

Security audit verdict: SAFE · quality 79/100

No red flags found in any of the 11 categories — no credential harvesting, no data exfiltration, no curl-pipe-shell installer. Scanned against the SlowMist agent-security taxonomy, refreshed every 8 hours. Full audit →

by ghost-in-the-droid · MCP Server · ★ 308

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

🔒 Is android-agent safe to install? View the security audit →

About android-agent

Ghost in the Droid Summon a ghost into your Android. It sees the screen. It taps the buttons. It never sleeps. Website · Docs · Skill Hub What It Does Open-source Python framework for controlling real Android phones via ADB. No app installed. No footprint. Pure automation. Define skills for any app, run them from the dashboard or API, scale across a phone farm. The ghost taps what you'd tap 50+ ADB methods — tap, swipe, type, clipboard, stealth variants Live phone screen streaming (MJPEG and WebRTC) Interactive touch-to-tap on the streamed screen Multi-device phone farm with per-device job queues Forge reusable skills for any app YAML-based UI element definitions per app Python action classes with precondition checks Multi-step workflows that chain actions together Built-in skills for TikTok and Play Store Skill Hub — browse, search, and install skills from the community r

adbandroidandroid-automationautomationbot-frameworkfastapiiosiphonemcpopen-source

Quick Facts

Stars308
Forks34
LanguagePython
CategoryMCP Server
LicenseMIT
Quality Score79.0745077137751/100
Open Issues1
Last Updated2026-08-13
Created2026-04-08
Platformsbrowser, mcp, python
Est. Tokens~16k

android-agent alternative? Top 6 similar tools

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

  • mobileClaw by eggbrid2 · ⭐ 410

    Open Android AI agent runtime for phone control, app automation, VLM screen reading, skill routing, mini apps,

  • flutter-skill by ai-dashboad · ⭐ 362

    AI-powered E2E testing for 10 platforms. 253 MCP tools. Zero config. Works with Claude, Cursor, Windsurf, Copi

  • dsh-ios by ZSeven-W · ⭐ 290

    DeepSeek Harness (DSH) plugin: a live iOS Simulator — and a USB-connected iPhone — inside the conversation. 22

  • OmoiOS by kivo360 · ⭐ 71

    Turn feature specs into merged PRs with a self-supervising swarm of coding agents — parallel execution, isolat

  • ios-simulator-skill by conorluddy · ⭐ 1.2k

    An IOS Simulator Skill for ClaudeCode. Use it to optimise Claude's ability to build, run and interact with you

  • cheatengine-mcp-bridge by miscusi-peek · ⭐ 1.2k

    Connect Cursor, Copilot & Claude AI directly to Cheat Engine via MCP. Automate reverse engineering, pointer sc

More MCP Server Tools

Explore other popular mcp server tools:

View all MCP Server tools →

Popular Python Agent Tools

Frequently Asked Questions

What is android-agent?

android-agent is Open-source framework to drive a real phone with AI agents: Android over ADB, iPhone over WebDriverAgent, 62 MCP tools, Python skills, a Vue dashboard, and swappable local or cloud LLMs. It is categorized as a MCP Server with 308 GitHub stars.

What programming language is android-agent written in?

android-agent is primarily written in Python. It covers topics such as adb, android, android-automation.

How do I install or use android-agent?

You can find installation instructions and usage details in the android-agent GitHub repository at github.com/ghost-in-the-droid/android-agent. The project has 308 stars and 34 forks, indicating an active community.

What license does android-agent use?

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

What are the best alternatives to android-agent?

The top alternatives to android-agent on Agent Skills Hub include mobileClaw, flutter-skill, dsh-ios. Each offers a different approach to the same problem space — compare them side-by-side by stars, quality score, and community activity.

How this security grade is produced

Grades come from a rule-based scan built on the SlowMist agent-security taxonomy, covering 11 red-flag categories including credential harvesting, data exfiltration, and curl | sh installers. It is a first-layer scan, not a manual audit — we say so rather than overstate it.

The scale of the problem is documented independently: Liu et al. (2026), in a study of 31,132 agent skills, report that 26.1% contain security vulnerabilities. Our own full-catalog census is published as a citable open dataset.

Sources & who's responsible:

View on GitHub → Browse MCP Server tools