Reverse engineered prompt

GitHub

Build me a first version of an AI workspace called Vincent AI Systems.

I want it to feel like an autonomous intelligence layer that can keep working through tasks in a loop, hold onto important context from a conversation, and organize that context in a useful way instead of losing the thread. It should also include a Grok deep exporter so I can pull conversation content or related outputs into this system and keep everything connected.

Please make it usable as a real foundation, not just a demo. I should be able to start it, feed it conversation context, let it process or iterate on tasks, and review what it saved, inferred, or exported. If the zip already contains most of the code, wire it up into one coherent project and fill in whatever is missing.

Keep the setup simple and document how to run it. If anything is unclear from the files, check current docs online and make sensible choices.

Have a live product UI? Try website reverse