toyxyz/ComfyUI_toyxyz_test_nodes ? reverse-engineered prompt

Build me a ComfyUI custom nodes pack that brings together the everyday tools I use for image and video workflows.

I want it to include things like regional prompting with masks, an easy way to draw or define multiple masked areas, an OpenPose editor with pose interpolation, webcam and screen capture, render preview and image saving tools, video export, and a few utility nodes for things like depth to normal, normal cleanup, random text loading, and glb export. It should feel like a practical toolbox for doing iterative AI art work faster, not a huge complicated app.

If you can, make the nodes easy to drop into a ComfyUI custom_nodes folder, and include a couple of example workflows so I can see how the pieces connect. Also add a newer prompt building node for audiovisual generation if that fits, where I can enter shots, attach image, video, or audio references, and generate a structured prompt from them. If you need current docs or APIs, look them up online first.