Wildflower Child

Wildflower Child is a small generative toy built around botanical matchmaking. Drag two flowers together and, when the board has room, their traits combine into a new bloom. There is no score to chase—just an antique-looking garden that gets stranger and fuller as you rearrange it.

What is Wildflower Child?

The rules fit comfortably on a seed packet: move a flower beside another, leave an open space nearby, and watch a child appear with a mixture of its parents' genomes. That simple adjacency rule turns the board into both sketchbook and family tree. Brad Buchanan made the project for GitHub Game Off 2017, using the jam's Throwback theme as a cue for its antique-illustration look. The source remains public under the MIT License, and a post-jam update added touch handling for mobile devices.

What you can do there

• Drag flowers around a board
• Pair neighboring flowers to create offspring
• Fill the board with new combinations

Why we picked it

Generative toys are often buried under sliders and control panels. This one makes experimentation feel physical: shuffle the garden, introduce a new pairing, and inspect the offspring. The lack of an objective is a feature, not an omission, because the pleasure comes from spotting how one small arrangement changes the next crop of flowers.

How to get the most from it

Drag a flower next to a second flower while leaving at least one neighboring space open. A child bloom can then occupy that gap with traits mixed from both parents. Keep rearranging the board to encourage new pairings, or stop as soon as you have made a tiny bouquet worth admiring.

Good to know

The toy is free, runs in the browser, and lists both mouse and touchscreen input. It has no stated win condition and can be sampled in seconds or fussed over much longer. Current mobile compatibility, keyboard navigation, screen-reader behavior, audio, and the sharing workflow were not verified during this review.

Who made it, and when?

Brad Buchanan is the credited creator or organization. The earliest supported launch date we found is December 1, 2017.