Run Tasks in Autopilot. Any model. Minimal setup.

Features · Quick Start · Demos

🎯 Overview

BrowserMind is a browser automation agent built on top of browser-use. giving you a single interface to run any browser task using the LLM of your choice. Point it at a task, pick a model, and let it run.

Form automation · Data scraping · Complex workflows · Automation

Task = "Fill in this job application with my resume and information."

✨ Features

🚀 Quick Start

1. Clone & Install

git clone https://github.com/rojansapkota/BrowserMind.git
cd BrowserMind
pip install -r requirements.txt

2. Set Your API Key

Create a .env file in the project root:

GROQ_API_KEY=your_groq_api_key
ANTHROPIC_API_KEY=your_anthropic_api_key
OPENAI_API_KEY=your_openai_api_key
GOOGLE_API_KEY=your_google_api_key

3. Run Your First Task

python main.py

When prompted, enter a task:

Enter your task for the agent: Find the latest AI news on Hacker News

The agent will automatically browse and complete the task for you.

💾 Installation

Prerequisites

• Python 3.11.9

Step-by-Step Setup

# 1. Clone the repository
git clone https://github.com/rojansapkota/BrowserMind.git
cd BrowserMind

# 2. Install dependencies
pip install -r requirements.txt

# 3. Install Chromium (if not already installed)
# Browser-Use handles this automatically, but you can manually install:
# python -m playwright install chromium

# 4. Set up your environment variables
cp .env.example .env  # Create from template (optional)
# Edit .env with your API keys

Demos

🍎 Grocery-Shopping

Task = "Put this list of items into my instacart."

https://github.com/user-attachments/assets/a6813fa7-4a7c-40a6-b4aa-382bf88b1850

💻 Personal-Assistant.

Task = "Help me find parts for a custom PC."

https://github.com/user-attachments/assets/ac34f75c-057a-43ef-ad06-5b2c9d42bf06

⚙️ Configuration

Supported LLM Providers

Environment Variables

# API Keys
GROQ_API_KEY=your_key_here
ANTHROPIC_API_KEY=your_key_here
OPENAI_API_KEY=your_key_here
GOOGLE_API_KEY=your_key_here

# Ollama (for local models)
OLLAMA_BASE_URL=http://localhost:11434

Customization Options

Edit main.py to customize:

# Browser window size
window_size={'width': 1280, 'height': 720}

# Page load wait time
minimum_wait_page_load_time=0.1

# Wait between actions
wait_between_actions=0.1

# Temperature (0 = deterministic, higher = more creative)
temperature=0.0

📖 Usage Examples

Basic Task Execution

from main import run_agent
import asyncio

# Simple task
asyncio.run(run_agent("Find the current Bitcoin price", provider='groq'))

# Use different providers
asyncio.run(run_agent("Fill out this form", provider='anthropic'))
asyncio.run(run_agent("Scrape product listings", provider='openai'))

Programmatic Usage

import asyncio
from main import initialize_agent

async def automated_workflow():
    query = "Navigate to google.com and search for 'Browser Mind'"
    agent, browser_session = initialize_agent(query, provider='groq')

    try:
        result = await agent.run()
        print(f"Task completed: {result}")
    finally:
        await browser_session.close()

asyncio.run(automated_workflow())

Common Use Cases

• Form Filling: Automatically complete and submit web forms
• Data Scraping: Extract structured data from websites
• Research: Gather information from multiple sources
• Monitoring: Watch for changes and alerts
• Automation: Repetitive browser tasks

🔧 Advanced Configuration

Custom System Message

CUSTOM_PROMPT = """
Your custom instructions here.
Focus on speed and accuracy.
"""

agent = Agent(
    task=query,
    llm=llm,
    browser_session=browser_session,
    extend_system_message=CUSTOM_PROMPT
)

Using Different Ollama Models

# Pull a model first
ollama pull mistral

# Then use it in main.py
return ChatOllama(model='mistral')

📊 Troubleshooting

🎓 Learn More

• Browser-Use Documentation: docs.browser-use.com
• Browser-Use GitHub: github.com/browser-use/browser-use
• Playwright Documentation: playwright.dev

🤝 Contributing

Contributions are welcome! Please feel free to:

• Report bugs via Issues
• Submit pull requests with improvements
• Share your use cases and examples

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.