Osmantic/ODS ? reverse-engineered prompt

Build me a simple way to turn a regular PC, Mac, or Linux machine into a private AI server that runs mostly on local hardware. I want a one step install experience, then I should be able to open a browser on localhost and chat with local models right away. It should also include a clean control panel to manage services, models, setup, ports, and basic system or GPU status.

Please wire in the main things people expect from a home AI setup, like chat, voice input and output, document search and RAG, automations and agents, and local image generation. Keep privacy front and center so prompts and files stay on the machine by default. If someone has no GPU, make sure there is an optional cloud mode that still uses the same interface.

Make it work across Linux, Windows, and Apple Silicon Macs as described in the repo, and keep the install and first run experience friendly for non technical users. If you need details, check the current docs online and follow the repo README.