Web Automation Bot

An intelligent Python-based web automation bot supporting both traditional programming and AI-driven task execution.

Features

Core Functionality (Required Core)

• Playwright-based browser automation
• Complete error handling and reliability guarantees
• Clear console output and logging
• Support for multiple browser operations (click, type, wait, screenshot, etc.)

AI Brain (Optional Challenge 1)

• Model Context Protocol (MCP) integration
• OpenAI GPT-4-driven dynamic task planning
• Intelligent page analysis and element identification
• Natural language goal to automation steps conversion

API Service (Optional Challenge 2)

• FastAPI-based RESTful API
• Support for remote execution of traditional and AI tasks
• Complete request/response model validation
• CORS support and health checks

Quick Start

1. Environment Setup

# Clone the project
cd web-automation-bot

# Install dependencies
pip install -r requirements.txt

# Install Playwright browser
playwright install chromium

2. Environment Configuration

Create a .env file for AI features (optional):

# OpenAI Configuration
OPENAI_API_KEY=your_openai_api_key_here

# MCP Server Configuration
MCP_SERVER_URL=http://localhost:3000

# Browser Configuration
BROWSER_HEADLESS=true
BROWSER_TIMEOUT=30000

# API Configuration
API_HOST=0.0.0.0
API_PORT=8000

# Logging Configuration
LOG_LEVEL=INFO

3. Run Examples

Core Functionality Example

python examples/simple_task.py

AI Brain Example (requires OpenAI API Key)

python examples/ai_task.py

API Service

python api/main.py

API Test (API service must be running)

python examples/api_example.py

API Documentation

Once the API service is running, visit http://localhost:8000/docs for complete API documentation.

Main Endpoints

• GET / - API information
• GET /health - Health check
• POST /execute-task - Execute predefined task
• POST /execute-ai-task - Execute AI-driven task

Request Examples

Execute Simple Task

{
  "url": "https://example.com",
  "steps": [
    {
      "action": "wait",
      "selector": "input[name='search']",
      "timeout": 10000,
      "description": "Wait for search box"
    },
    {
      "action": "type",
      "selector": "input[name='search']",
      "text": "Python automation",
      "description": "Type search term"
    },
    {
      "action": "click",
      "selector": "button[type='submit']",
      "description": "Click submit button"
    }
  ]
}

Execute AI Task

{
  "goal": "Get page title",
  "url": "https://example.com"
}

Project Structure

web-automation-bot/
├── README.md                 # Project documentation
├── requirements.txt          # Dependencies
├── config.py                 # Configuration management
├── core/                     # Core functionality
│   ├── __init__.py
│   ├── browser_driver.py     # Browser driver
│   └── task_executor.py      # Task executor
├── ai_brain/                 # AI brain module
│   ├── __init__.py
│   ├── mcp_client.py         # MCP client
│   ├── llm_handler.py        # LLM handler
│   └── task_planner.py       # Task planner
├── api/                      # API service module
│   ├── __init__.py
│   ├── main.py               # FastAPI application
│   └── models.py             # Data models
├── examples/                 # Usage examples
│   ├── simple_task.py        # Simple task example
│   ├── ai_task.py            # AI task example
│   └── api_example.py        # API usage example
└── tests/                    # Test files
    ├── __init__.py
    ├── test_core.py          # Core functionality tests
    └── test_ai_brain.py      # AI brain tests

Technology Stack

• Python 3.8+ - Main programming language
• Playwright - Browser automation
• FastAPI - Web API framework
• OpenAI GPT-4 - AI model
• Pydantic - Data validation
• httpx - HTTP client
• pytest - Testing framework

Extending the Project

Adding Browser Operations

Edit core/browser_driver.py:

async def new_action(self, param: str) -> bool:
    """Add new browser action"""
    try:
        # Implement your action
        return True
    except Exception as e:
        logger.error(f"Action failed: {e}")
        return False

AI Features

Edit ai_brain/llm_handler.py to customize AI behavior.

API Endpoints

Edit api/main.py to add new endpoints:

@app.post("/new-endpoint")
async def new_endpoint(request: NewRequestModel):
    # Implement your logic
    pass

Running Tests

# Run all tests
pytest

# Run specific test files
pytest tests/test_core.py
pytest tests/test_ai_brain.py