Token导航 LogoToken导航TokenDH.com
研究检索执行命令github未标认证来源可访问许可证需确认审计异常

burner-phone一次性电话

Agent Skill

burner-phone 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

总安装

285

周安装

12

GitHub Stars

20

下载量

1
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

unknown

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:burner-phone(一次性电话)
来源仓库:https://github.com/southpawin/burner-phone
仓库路径:skills/burner-phone
安装命令:
npx skills add https://github.com/southpawin/burner-phone --skill burner-phone
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/southpawin/burner-phone --skill burner-phone

简介

burner-phone 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 中需要根据关键词快速定位候选结果时使用。

  • 它通过 Python API 控制物理设备,支持拍照、录音和屏幕操作。
  • 使用时需配置设备类型和连接参数,测试前确保硬件就绪。
  • 安装前应确认是否会触发摄像头或麦克风访问,遵守隐私法规。
  • 该技能输出为设备指令,需人工监控执行过程。

SKILL.md

🔥 Universal Phone Agent Control

Quick Start (v2.0)

Setup

# 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)

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

Step 2: Analyze with Vision

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

Step 3: Act with Exact Coordinates

# 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(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

# 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

# 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

# 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

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

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)

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

Emulator

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

Advanced Usage

Multi-Device Management

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

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"

# 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"

# 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

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Codex

37.88%
按下载量换算0

Claude

31.16%
按下载量换算0

Cursor

18.52%
按下载量换算0

Gemini CLI

8.83%
按下载量换算0

安全审计

Gen Agent Trust Hub

通过

Socket

可疑

Snyk

未通过

权限和风险

执行命令

安装流程涉及命令执行,可能通过 npx skills add https://github.com/southpawin/burner-phone --skill burner-phone 联网下载 Skill 或依赖。用户安装前应确认命令来源、仓库内容和执行环境。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。来源安全扫描存在 warning/failed 结果,不能写成本站确认安全。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills