It’s an agent-ready image-to-font extractor: image in, known character order in, TTF/SVG/manifest/contact sheet/preview out.