Burner Phone logo

Burner Phone

Community
SouthpawIN
burner-phone

Universal Android device control with vision feedback. Supports Termux phones, ADB-only devices, and emulators. Use for phone automation, AI companionship, or mobile app testing.

Overview

PublisherSouthpawIN
Repositoryburner-phone
Skill nameburner-phone
Stars
58
Forks
7
Bundled files
42
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 42 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by SouthpawIN on GitHub. Read the source before you install it.

Installation

Install the Burner Phone AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/SouthpawIN/burner-phone/tree/main
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/SouthpawIN/burner-phone.git \
  .claude/skills/burner-phone
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Burner Phone in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Burner Phone on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Burner Phone is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

🔥 Universal Phone Agent Control

Quick Start (v2.0)

Setup

bash
# Configure your device first
cp config.example.yaml ~/.hermes-phone-agent/config.yaml
# Edit the config file for your device type

# Test connection
python3 scripts/test_device.py

Python API (New in v2.0)

python
from phone_agent import PhoneAgent

# Auto-detects device from config
agent = PhoneAgent()

# Camera
agent.capture_camera()  # Captures to configured path

# Audio
agent.record_audio(duration=5, output_path="recording.wav")
agent.play_audio("response.wav")

# Screen control
agent.wake_screen()
agent.unlock_screen(pin="4658")
agent.type_text("Hello world")

# Check device status
print(agent.device.is_online())

Vision Feedback Loop (v1 Compatible)

ALWAYS follow this pattern for UI automation:

Step 1: Screenshot

bash
bash(cmd="adb exec-out screencap -p > ./assets/screen.png")

Step 2: Analyze with Vision

bash
bash(cmd="python3 ./scripts/vision_helper.py ./assets/screen.png \"Find all buttons and their coordinates (x,y)\"")

Step 3: Act with Exact Coordinates

bash
# Tap at coordinates from vision analysis
bash(cmd="adb shell input tap <x> <y>")

# Swipe
bash(cmd="adb shell input swipe <x1> <y1> <x2> <y2> 300")

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

Step 4: Verify

bash
bash(cmd="adb exec-out screencap -p > ./assets/screen_after.png")
bash(cmd="python3 ./scripts/vision_helper.py ./assets/screen_after.png \"Confirm action succeeded\"")

Device Types Supported

Termux (Best Performance)

  • Physical Android phones with Termux installed
  • Direct camera access via termux-camera-photo
  • Audio recording via termux-microphone-record
  • Playback via termux-media-player
  • SSH + SCP for file transfers

ADB Only (Standard Android)

  • Any Android device with ADB enabled
  • Camera via screencap (screen capture)
  • Audio via screenrecord --audio (Android 10+)
  • Push/pull files via ADB

Emulator (No Hardware Needed)

  • Android Studio emulator
  • Genymotion
  • Waydroid (Linux container)
  • Outputs audio to host speakers

Available Commands

Basic Actions

bash
# Tap at coordinates
bash(cmd="adb shell input tap <x> <y>")

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

# Type text (escapes special chars)
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")
bash(cmd="adb shell input keyevent KEYCODE_WAKEUP")

App Control

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

# Force stop app
bash(cmd="adb shell am force-stop com.package.name")

# Open URL
bash(cmd="adb shell am start -a android.intent.action.VIEW -d 'https://example.com'")

Screen Management

bash
# Wake screen
bash(cmd="adb shell input keyevent KEYCODE_WAKEUP")

# Take screenshot
bash(cmd="adb exec-out screencap -p > ./assets/screen.png")

# Record screen with audio (Android 10+)
bash(cmd="adb shell screenrecord --audio /sdcard/recording.mp4")

Configuration Examples

Samsung Galaxy S10 with Termux

yaml
name: "Samsung Galaxy S10"
device_type: "termux"
ip_address: "100.93.96.90"
ssh_port: 8022
ssh_key: "~/.ssh/phone_access"
screen_pin: "4658"

Surface Duo 2

yaml
name: "Surface Duo 2"
device_type: "termux"
ip_address: "100.79.15.54"
ssh_port: 8022
ssh_key: "~/.ssh/phone_access"

Standard Android (No Termux)

yaml
name: "Pixel Phone"
device_type: "adb"
ip_address: "100.x.x.x"
screen_pin: "1234"

Emulator

yaml
name: "Android Studio Emulator"
device_type: "emulator"
ip_address: "localhost"
adb_port: 5555
screen_pin: null

Advanced Usage

Multi-Device Management

python
from phone_agent import PhoneAgent
from config.device_config import load_config

# Load different configs
config1 = load_config("~/.hermes-phone-agent/phone1.yaml")
config2 = load_config("~/.hermes-phone-agent/phone2.yaml")

# Create agents for each
agent1 = PhoneAgent(config1)
agent2 = PhoneAgent(config2)

# Control both devices
agent1.type_text("Hello from phone 1")
agent2.type_text("Hello from phone 2")

Vision-Driven Automation

python
from phone_agent import PhoneAgent
import requests

agent = PhoneAgent()

# Screenshot
agent.device._adb_shell("screencap -p /sdcard/screen.jpg")
agent.device._adb_pull("/sdcard/screen.jpg", "./assets/screen.png")

# Send to vision API (using Senter Server)
response = requests.post(
    "http://localhost:8081/v1/chat/completions",
    json={
        "model": "qwen2.5-omni:3b",
        "messages": [{
            "role": "user",
            "content": [
                {"type": "text", "text": "Find the Settings button coordinates"},
                {"type": "image_url", "image_url": {"url": "data:image/png;base64,..."}}
            ]
        }]
    }
)

# Parse coordinates and tap
# (implement coordinate parsing from vision response)
x, y = parse_coordinates(response.json())
agent.device._adb_shell(f"input tap {x} {y}")

Rules & Best Practices

  1. ALWAYS screenshot before acting - Never guess coordinates
  2. ALWAYS use vision_helper.py to get precise element locations
  3. Use coordinates EXACTLY as provided by vision analysis
  4. Verify after each action - Screenshot again to confirm
  5. Handle device state - Wake and unlock screen before operations
  6. Respect timeouts - Give devices time to respond
  7. Test on emulators first - Safer than physical devices

Troubleshooting

"Device not found"

bash
# Check ADB connection
adb devices

# For Termux, verify SSH
ssh -i ~/.ssh/phone_access -p 8022 droid@<IP>

# Reconnect if needed
adb connect <IP>:5555

"Permission denied"

bash
# On physical device, enable USB debugging and authorize computer
# For wireless ADB, enable "Wireless debugging" in Developer Options

"Camera not working"

  • Termux devices: Install termux-camera-photo via pkg
  • ADB-only: Uses screencap (shows current screen, not direct camera)
  • Emulators: Virtual camera may need configuration

Changelog

v2.0 (Current)

  • ✅ Universal device abstraction layer
  • ✅ Termux, ADB, and Emulator backends
  • ✅ Configuration-driven setup
  • ✅ Python API for integration
  • ✅ Multi-device management
  • ✅ Backward compatible with v1

v1.0

  • ✅ Vision feedback loop
  • ✅ ADB control commands
  • ✅ Openskills format support

Related

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Burner Phone AI skill do?

Universal Android device control with vision feedback. Supports Termux phones, ADB-only devices, and emulators. Use for phone automation, AI companionship, or mobile app testing.

Why use Burner Phone on TypingMind?

Because you install it once and use it with any model. Burner Phone is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Burner Phone in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/SouthpawIN/burner-phone/tree/main. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Burner Phone?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Burner Phone?

As many as you like. As long as a model supports skills, you can use Burner Phone with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Burner Phone AI skill free?

Yes. It is published on GitHub by SouthpawIN under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇