Omnidim/omnidim-python-sdk ? reverse-engineered prompt

Build me a Python SDK for OmniDimension that makes it easy to create and manage AI voice agents from code.

I want to be able to set my API key, connect to the service, list agents, explore available AI providers like voices and LLMs, attach knowledge base files, add integrations, manage phone numbers, and launch bulk calling campaigns. Please include simple examples I can run right away, and make the SDK feel clean and easy to use for someone who just wants to automate voice agent setup without dealing with a lot of complexity.

Also add a small example script that shows the main flow, and make sure the package is ready to install and use in Python 3.9 or newer. If anything in the API is unclear, look up the current docs online and mirror the real behavior as closely as you can.