OpenAtlas Agent

An autonomous browser agent with web search and interactive browsing capabilities. OpenAtlas can search the web using DuckDuckGo, browse websites interactively, click elements, fill forms, and extract information

OpenAtlas Agent has long term memory (local)

Note: This project requires Node.js support. Make sure you have Node.js installed on your system.

Installation

pip3 install openatlas

API Keys Setup

OpenAtlas requires an API key for your chosen LLM provider. Add one of the following to your environment:

Anthropic (Claude)

export ANTHROPIC_API_KEY=your_key_here

OpenAI (GPT)

export OPENAI_API_KEY=your_key_here

Google (Gemini)

export GEMINI_API_KEY=your_key_here

Any LLM supported by LiteLLM

OpenAtlas supports 100+ models from various providers including:

• OpenAI (GPT-4, GPT-3.5)
• Anthropic (Claude)
• Google (Gemini)
• Cohere
• Mistral AI
• And many more!

See LiteLLM documentation for the full list of supported models.

Quick Start

# Start OpenAtlas Agent
openatlas

# Or run directly with Python
python cli.py

# Specify a model
python cli.py --model gpt-4o

Usage

Commands

All commands start with /:

• /model - Choose a different AI model
• /status - Show current session configuration
• /clear - Clear the screen
• /help - Show detailed help information
• /quit - Exit the CLI

Example Tasks

Web Search

> What's happening in AI today?
> Search for Python 3.13 new features
> Find the latest news about SpaceX

Website Browsing

> Open https://news.ycombinator.com and list the top stories
> Browse to github.com and search for AI projects
> Go to wikipedia.org and find information about Mars

Interactive Tasks

> Open reddit.com/r/programming and show me the trending posts
> Search Google for "best restaurants near me"
> Navigate to amazon.com and search for wireless headphones

Research Tasks

> Research on multiple websites and find top whey protein brands in 2025
> Compare prices for MacBook Pro across different retailers
> Find reviews for the latest iPhone from tech websites
> Browse 5 different tech news sites and summarize today's biggest story

Use Cases

🔬 Research & Information Gathering

• Multi-source research: Gather information from multiple websites simultaneously
• Competitive analysis: Research competitors' products, pricing, and strategies
• Academic research: Collect data and citations from various web sources
• Product comparison: Compare features, prices, and reviews across platforms

Example:

> Research on multiple websites and find top whey protein brands in 2025

📰 News & Updates

• Breaking news monitoring: Track news from multiple sources in real-time
• Topic tracking: Follow specific topics across different news outlets
• Social media aggregation: Gather trending discussions from social platforms
• Industry trends: Stay updated with the latest developments in your field

Example:

> Open https://news.ycombinator.com and list the top stories
> Check latest AI news from TechCrunch and The Verge

🛍️ Shopping & E-commerce

• Product research: Find and compare products across retailers
• Price comparison: Track prices across different e-commerce platforms
• Review aggregation: Collect and analyze product reviews
• Availability checking: Check stock across multiple stores

Example:

> Compare prices for iPhone 15 Pro on Amazon, Best Buy, and Apple Store
> Find the best-rated wireless earbuds under $100

💼 Business Intelligence

• Competitor monitoring: Track competitor websites and announcements
• Market trend analysis: Analyze market trends from multiple sources
• Lead generation: Find potential customers and partners online
• Industry news tracking: Monitor industry-specific news and updates

Example:

> Monitor tech startup funding announcements from TechCrunch this week
> Research top SaaS companies in the project management space

🎓 Learning & Education

• Research for assignments: Gather information for academic projects
• Finding educational resources: Discover tutorials, courses, and documentation
• Staying updated: Learn about latest developments in any field
• Exploring topics: Deep dive into subjects with guided research

Example:

> Find the official Python documentation and explain the new features in 3.13
> Research the history of artificial intelligence from multiple sources

🎯 Content Creation

• Content research: Gather information for articles and blog posts
• Fact-checking: Verify information across multiple sources
• Trend analysis: Identify trending topics and discussions
• Source collection: Compile references and citations

Example:

> Research the top 10 AI trends for 2025 from multiple tech blogs
> Find statistics about remote work adoption in 2024

Supported Models

Default Models

• claude-sonnet-4-5 (Anthropic) - Default, best for complex tasks
• gpt-4o (OpenAI) - Excellent for general tasks
• gemini-2.0-flash (Google) - Fast and efficient

Custom Models

Use any model supported by LiteLLM by selecting option 4 during model selection or using the /model command. This includes:

• gpt-4, gpt-3.5-turbo (OpenAI)
• claude-3-opus, claude-3-sonnet (Anthropic)
• gemini-pro, gemini-1.5-pro (Google)
• command-r-plus (Cohere)
• mistral-large (Mistral AI)
• And 100+ more models!

License

MIT License

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

Support

For issues, questions, or feature requests, please open an issue on GitHub.

---

Made with ❤️ by Harish Santhanalakshmi Ganesan