MinionWorks – Modular browser agents that work for bananas 🍌

Modular. Extensible. AI-native browser agents for modern web automation.

---

🚀 Overview

MinionWorks is a modular AI agent framework that connects to your browser and executes complex tasks autonomously. Built for developers, researchers, and curious builders.

✨ Features

• 🌐 Perform Google searches and scrape content
• 🤖 Use LLMs (like GPT-4) to plan actions
• 🔗 Modular architecture for plug-and-play use cases
• 🔎 DOM interaction & content extraction
• 🔄 Run workflows via Python or UI

---

🛠️ Installation

1. Install the package

   pip install minion-agent
   

2. Set up environment variables

   cp .env.example .env
   # Edit your .env file with OpenAI or other API keys
   

---

🧪 Quick Start

Here’s a complete example using MinionAgent with langchain-openai:

from langchain_openai import ChatOpenAI
from minion_agent.browser import MinionAgent
import asyncio
from dotenv import load_dotenv

# Load environment variables from .env file
load_dotenv()

async def main():
    # Example 1: Using environment variables for OpenAI configuration
    agent1 = MinionAgent(
        task="Compare the price of GPT-4 and DeepSeek-V3",
        llm=ChatOpenAI(model="gpt-4o"),
        headless=True
    )
    result1 = await agent1.run()
    print("Result 1:", result1)

if __name__ == "__main__":
    asyncio.run(main())

---

💡 Example Use Cases

Invoke MinionAgent with any of these example prompts to see it in action:

1. List of Indian B2B YC Companies

   agent = MinionAgent(task="Give me the list of all B2B YC companies in India", llm=ChatOpenAI(model="gpt-4o"))
   result = await agent.run()
   print(result)
   

2. Kerala CPI Index (Feb 2024)

   agent = MinionAgent(task="What is the CPI index of kerela for the month of february 2024", llm=ChatOpenAI(model="gpt-4o"))
   result = await agent.run()
   print(result)
   

3. 3-Day Ladakh Itinerary from Delhi

   agent = MinionAgent(task="Give me the itinerary for the three day plan for visiting Ladakh from Delhi", llm=ChatOpenAI(model="gpt-4o"))
   result = await agent.run()
   print(result)
   

---

📂 Examples & Recordings

Play back the recorded demos in the examples/videos/ directory and see MinionAgent tackle each prompt below.

1. Indian B2B YC Companies

Prompt:

Give me the list of all B2B YC companies in India

---

2. Kerala CPI Index (Feb 2024)

Prompt:

What is the CPI index of kerela for the month of february 2024

---

3. Ladakh Itinerary from Delhi

Prompt:

Give me the itinerary for the three day plan for visiting Ladakh from Delhi

---

🧪 Testing

pytest --maxfail=1 --disable-warnings -q

Ensure you’re in the root folder where tests/ lives.

---

🤝 Contributing

We welcome PRs, feedback, and creative ideas!

1. Fork → Branch → Commit
2. Add tests
3. Submit a Pull Request
4. Tell your friends 🚀

---

📖 Citation

@software{minion_works2025,
  author = {Sairaam, Aman, Cheena},
  title = {Minion Works: Let AI take the helm of your browser.},
  year = {2025},
  publisher = {GitHub},
  url = {https://github.com/minionworks/minions}
}