MaxLiu22/Motion2MixamoRig ? reverse-engineered prompt

Build me a tool that takes a short video of one person, or a single full body photo, and transfers that motion or pose onto a Mixamo rigged 3D character.

I want a simple command line flow where I can first check that my files are in the right place, then run the pipeline and get output files I can use in Blender or Unity. It should accept a video or an image, use a Mixamo character FBX, and create preview videos or stills, plus a skinned .glb and motion data files. If I run it again with the same extracted skeleton, it should let me reuse that so the second pass is much faster.

Please include clear messages for missing assets, a default Mixamo character if one is present, and sensible output folders named by run time. If there is a Blender add on version too, make sure it can launch the same pipeline from Blender. Look up current docs online if you need to.