Robbyant/lingbot-world ? reverse-engineered prompt

Build me a simple working demo for LingBot World so I can turn an input image and text prompt into a generated video, with optional camera pose or action control if I have it.

I want the project to be easy to run from the command line, and if possible, also give me a small local web page where I can upload an image, type a prompt, choose the model, and start generation. Keep it focused on inference, not training.

Use the existing scripts and examples in the repo as the starting point, and make sure it works for both the normal and fast generation modes. If there are any missing setup details or model download steps, please add clear instructions in the README. Look up current docs online if you need to.