What Pagenator is

Pagenator is a reconstruction spec tool, not a page builder. Your uploaded design is a reference, never the final graphic. You mark each visual part, say what it is, and Pagenator decides how it should be rebuilt — as code, as vector, or as a raster fallback.

design image → classified parts → component definitions → Lovable rebuilds the page

What the MVP does

· One uploaded design per scene, stored in your browser.
· Rectangular parts you draw, move, resize and label.
· Part classification: button, nav item, card, panel, input, text, icon, frame, separator, HUD element, ornament, container or texture.
· An output format per part — HTML/CSS component, SVG, or raster fallback for complex decorative surfaces only.
· One trigger and one effect per interactive part.
· An export package: brief, scene JSON, component definitions, reconstruction rules and three phased prompts.

What it deliberately isn't

· Never uses your screenshot as the finished page graphic.
· Never slices the design into a mosaic of image crops.
· Never bakes text into images — text is rebuilt as real text.
· No accounts, payments, or server storage.
· No autonomous AI reconstruction in v1; you confirm the parts yourself.

Boxes are stored as fractions of the design, so they act as proportional layout hints — the rebuilt page stays fluid and responsive rather than pixel-pinned to the reference.