spongecake 0.1.1

Spongecake is the easiest way to launch OpenAI computer use agents.

Open source SDK to launch OpenAI computer use agents 🤖

(coming soonsies) Spinning up a computer use agent with sponge box

What is spongecake?

🚀 spongecake is the easiest way to launch OpenAI computer use agents.

Prerequisites

You’ll need the following to get started (click to download):

• Docker
• OpenAI API Key (sign up at OpenAI)

Note: The versions mentioned above are tested references. Other versions may work, but haven’t been fully validated.

Quick Start

Install the spongecake package using pip:

pip install spongecake

Clone the repo and run test.py...

cd spongecake/test
python3 test.py

Done! Edit the script in test.py to test out sponge bob

Usage Details

[coming soon]

How it works

[coming soon]

Appendix

Contributing

Feel free to open issues for any feature requests or if you encounter any bugs! We love and appreciate contributions of all forms.

Pull Request Guidelines

1. Fork the repo and create a new branch from main.
2. Commit changes with clear and descriptive messages.
3. Include tests, if possible. If adding a feature or fixing a bug, please include or update the relevant tests.
4. Open a Pull Request with a clear title and description explaining your work.

Roadmap

[coming soon]

Team

Made with ❤️ in San Francisco

Backend set-up

• Set-up .env in the backend folder with OpenAI Key and Port

SDK set-up

• cd sdk
• (in spongecake) pip3 install --upgrade pip setuptools wheel build
• (in spongecake) pip install --upgrade openai # Make sure you get the latest responses API
• (in spongecake) python3.11 -m build
• (in spongecake) pip3 install dist/spongecake-0.1.2-py3-none-any.whl
• (in root) pip3 install -e spongecake/
• pip3 install dotenv

Testing the app

• python3 test.py

Building and running the docker container

• cd sdk/docker
• docker build -t local-lxde-vm
• docker run -d -p 5900:5900 --name myvm local-lxde-vm
• docker exec -it myvm bash

Connecting to virtual desktop

• Open TigerVNC viewer
• Connect to localhost:5900