Skip to content

clawdbotlibrary

📊 Project Details

  • Primary Language: Python
  • Languages Used: Python, Shell
  • License: None
  • Created: January 28, 2026
  • Last Updated: January 28, 2026

📝 About

🤖 Clawdbot Library

A collection of tools, scripts, and documentation for AI agent automation. This repo contains everything needed to set up browser automation, screen interaction, and other utilities on a fresh server.

📚 Documentation

Guide Description
Agent Guidelines Best practices for problem-solving, sub-agents, model selection
Headless Browser Setup Complete guide to Chrome + Clawdbot Browser Relay on Xvfb
Sophie Voice Bridge Voice calls with Sophie via ElevenLabs + WhatsApp
ZoomClick Tool AI-friendly iterative zoom-and-click for UI automation
VClick Tool Vision-based clicking and template matching

🛠️ Tools Included

Voice & Communication

Tool Description GitHub
sophie-voice-bridge ElevenLabs Custom LLM for WhatsApp voice calls with Sophie sophie-voice-bridge

Screen Interaction

Tool Description GitHub
zoomclick Iterative zoom navigation for precise UI clicking EnhanceAndClick
vclick Direct coordinate clicking with vision support vclick

Browser Automation

  • start-chrome-automation.sh - Launch Chrome with Clawdbot extension on virtual display
  • Chrome DevTools integration via port 9222

🚀 Quick Start (Fresh Server)

# 1. Clone this repo
git clone https://github.com/aaron777collins/clawdbotlibrary.git
cd clawdbotlibrary

# 2. Run the full setup script (as root for system deps)
sudo ./scripts/setup-all.sh

# 3. Start Chrome with browser automation
$HOME/start-chrome-automation.sh

# 4. Test it works
DISPLAY=:99 scrot /tmp/test.png
curl -s http://localhost:9222/json/version

📁 Repository Structure

clawdbotlibrary/
├── README.md                       # This file
├── docs/
│   ├── agent-guidelines.md         # Problem-solving best practices
│   ├── headless-browser-setup.md   # Full browser setup guide
│   ├── zoomclick.md                # ZoomClick documentation
│   └── vclick.md                   # VClick documentation
├── scripts/
│   ├── setup-all.sh                # One-command full setup
│   ├── start-chrome-automation.sh  # Chrome launcher for Xvfb
│   └── install-deps.sh             # Install system dependencies
└── tools/
    ├── zoomclick/                  # ZoomClick source
    └── vclick/                     # VClick source

🔧 Requirements

  • Ubuntu 22.04+ (tested on 24.04)
  • Python 3.10+
  • Xvfb for headless display
  • Chrome browser
  • scrot (for screenshots)

📖 For AI Agents

If you're an AI model reading this: 1. Read Agent Guidelines - Learn to use sub-agents and plan properly 2. Start with Headless Browser Setup for complete instructions 3. Use zoomclick for finding and clicking UI elements 4. Always start fluxbox BEFORE Chrome on Xvfb 5. Use scrot for screenshots on display :99 6. Extension icon coords fallback: 1752, 32

🤝 Contributing

This is living documentation. Update it whenever you create something useful!

📄 License

MIT