burner-phone

by @SouthpawIN in Productivity

# Install this skill:

npx skills add SouthpawIN/burner-phone

Or install specific skill: npx add-skill https://github.com/SouthpawIN/burner-phone

# Description

Control Android devices via ADB with vision feedback. Use this to see the screen, take screenshots, analyze UI elements, and automate phone tasks.

# SKILL.md

name: burner-phone
description: Control Android devices via ADB with vision feedback. Use this to see the screen, take screenshots, analyze UI elements, and automate phone tasks.
model: qwen2.5-omni:3b
keywords: android, phone, adb, screenshot, vision, screen, tap, swipe, automation

Burner Phone Control

Use this skill for ANY request involving phone screens or mobile app automation.

Vision Feedback Loop

ALWAYS follow this pattern:

Screenshot: Capture the current screen
bash(cmd="adb exec-out screencap -p > ./assets/screen.png")
Analyze: Use vision model to understand the screen
bash(cmd="python3 ./scripts/vision_helper.py ./assets/screen.png \"Describe the screen and list coordinates (x,y) for interactable elements.\"")
Act: Perform the action using exact coordinates from step 2
bash(cmd="adb shell input tap <x> <y>")
Verify: Screenshot again to confirm the action worked

Available Commands

Tapping

bash(cmd="adb shell input tap <x> <y>")

Swiping

bash(cmd="adb shell input swipe <x1> <y1> <x2> <y2> <duration_ms>")

Typing Text

bash(cmd="adb shell input text 'your text here'")

Key Events

bash(cmd="adb shell input keyevent KEYCODE_HOME")
bash(cmd="adb shell input keyevent KEYCODE_BACK")
bash(cmd="adb shell input keyevent KEYCODE_ENTER")

Launch App

bash(cmd="adb shell am start -n com.package.name/.MainActivity")

Rules

ALWAYS screenshot before acting - never guess coordinates
ALWAYS use vision_helper.py to get coordinates
Use coordinates provided by the vision tool EXACTLY
All paths are relative to the skill root directory

# README.md

🔥 Burner Phone Skill

Burner Phone

Control Android devices directly via ADB commands - perfect for burner phones, testing devices, or automation tasks.

Features

Vision-First: Uses AI to analyze screen content and provide exact coordinates
Direct Control: ADB commands for tapping, swiping, typing
Openskills Compatible: Works with any agent that supports the openskills format

Quick Start

# Clone the skill to your skills directory
git clone https://github.com/SouthpawIN/burner-phone.git ~/.opencode/skills/burner-phone

# Run setup
cd ~/.opencode/skills/burner-phone
chmod +x scripts/setup.sh
./scripts/setup.sh

Requirements

Python 3.8+
ADB (Android Debug Bridge)

Configuration

Variable	Default	Description
`SENTER_URL`	`http://localhost:8081`	Senter Server URL
`VISION_MODEL`	`qwen2.5-omni:3b`	Vision model name

Usage

The skill follows a Vision Feedback Loop:

Screenshot → Capture current screen
Analyze → AI identifies UI elements and coordinates
Act → Execute ADB command with exact coordinates
Verify → Screenshot again to confirm

Example Commands

# Take screenshot
adb exec-out screencap -p > ./assets/screen.png

# Analyze screen
python3 ./scripts/vision_helper.py ./assets/screen.png "Find the Settings icon"

# Tap at coordinates
adb shell input tap 540 1200

# Swipe up
adb shell input swipe 540 1800 540 800 300

Directory Structure

burner-phone/
├── SKILL.md              # Skill manifest (openskills format)
├── README.md             # This file
├── scripts/
│   ├── vision_helper.py  # Vision analysis helper
│   └── setup.sh          # Installation script
└── assets/
    └── screen.png        # Screenshots saved here

Senter - Agent that uses this skill
Senter-Server - Model proxy server

License

MIT

# Supported AI Coding Agents

This skill is compatible with the SKILL.md standard and works with all major AI coding agents:

⚡ Amp 🚀 Antigravity 🤖 Claude Code 🦀 Clawdbot 📝 Codex ▶️ Cursor 🤖 Droid 💎 Gemini CLI 🐙 GitHub Copilot 🪿 Goose 📊 Kilo Code 🔧 Kiro CLI 💻 OpenCode 🦘 Roo Code 🌲 Trae 🏄 Windsurf

Learn more about the SKILL.md standard and how to use these skills with your preferred AI coding agent.