agex (a portmanteau of agent execution) is a Python-native agentic framework that enables AI agents to work directly with your existing libraries and codebase.

This works because agex agents can accept and return complex types like pandas.DataFrame and plotly.Figure objects without intermediate JSON serialization. For a deeper dive, check out the full agex101.ipynb tutorial or see geospatial routing with OSMnx for advanced multi-library integration.

For a full demo app where agex integrates with NiceGUI, see agex-ui.

agex uses a subset of Python as the agent action space, executing actions in a sandboxed environment within your process. This approach avoids the complexity of JSON serialization and allows complex objects to flow directly between your code and the agent. You control exactly what functions, classes, and modules are available, creating a safe and focused environment for the agent.

• Code-as-Action: Secure, sandboxed Python execution for agents.
• Library Integration: Use your existing code directly, no tool-making required.
• Workspace Persistence: Git-like versioning for agent state and memory.
• Multi-Agent: Orchestrate agents with natural Python control flow.
• Event Streams: Real-time, notebook-friendly observability.
• Benchmarking: A framework for data-driven agent evaluation.

Complete documentation is hosted at ashenfad.github.io/agex.

Key sections:

• 📚 Quick Start Guide
• 🔭 The Big Picture
• 💡 Examples
• 📖 API Reference

Install agex with your preferred LLM provider:

# Install with a specific provider
pip install "agex[openai]"        # For OpenAI models
pip install "agex[anthropic]"     # For Anthropic Claude models
pip install "agex[gemini]"        # For Google Gemini models

# Or install with all providers
pip install "agex[all-providers]"

⚠️ agex is a new framework in active development. While the core concepts are stabilizing, the API should be considered experimental and is subject to change.

For teams looking for a more battle-tested library built on the same "agents-that-think-in-code" philosophy, we highly recommend Hugging Face's excellent smolagents project. agex explores a different architectural path, focusing on deep runtime interoperability and a secure, sandboxed environment for direct integration with existing Python libraries.

We welcome contributions! See our Contributing Guide for details on our development workflow, code style, and how to submit pull requests. For bug reports and feature requests, please use GitHub Issues.