Hit "Generate README" and mkr.cards turned the spatial thinking into a proper spec.

Dropped the README into Lovable.

Built through 4 phases, mostly one-shot it. Minor polish after (gradient, borders, etc).

You can clone it and see the exact README I generated: germany2025.lovable.app

Started by mapping the spec out in mkr.cards

Described how I wanted:
- Photo stack layout
- Hover/drag physics
- Click interactions
- Card styling

Each card = a piece of functionality

Instead of letting my photos sit around on my camera, I built this: germany2025.lovable.app

Drag photos around and group them how you want. Simple but clean interactions.

Took ~1 hour from idea to working site.

New and improved card hover and pickup animations. Added a little wiggle when you pick up a card, smoother transitions, and a bunch of subtle improvements that make it feel more natural.

And... You can try it out yourself now at mkr.cards.

#buildinpublic #aitools #productdesign #aifordesigners

Step 4 looking good

Exactly what you wanted. Mission complete!

This is super cool! Maybe in addition to voice you could photograph your fridge and it could detect ingredients/suggest recipes too?

Originally specs were placed on the canvas, but it felt a bit clunky. The sidebar feels better but I'll be continuing to iterate here. I'll probably take this back to Figma and jam on some more layout options.

Also a sneak peek of github mcp so it can have context for feature updates!

Flowcharting usability updates and a new and improved canvas header, and toolbar!

Updated the toolbar by bringing over my component design in figma and used their MCP. It worked right away which was awesome.

Thank you! Yea I built it with Lovable. Sketched out a little README-style plan before starting, but the barbarian theme kinda just happened during some late-night vibe coding. Made the avatar in GPT and themed the site to match. Used tailwind for spacing, colors, rounded corners, etc.