Inspiration

Choosing a Halloween costume should be fun, not a last-minute panic. For a lot of girls, the hard part is not wanting to dress up, it is putting a whole look together that fits the occasion, the weather, and the budget, without buying the wrong pieces. Haunted Wardrobe Studio was built to solve that one night, with a plan you can actually wear.

What it does

Pick four things: occasion, vibe, budget, and comfort. The app returns one complete costume plan: main costume, accessories, hair and makeup, a DIY vs buy note, a cost breakdown that sums exactly to a total inside your budget band, and a tick-off checklist to wear and pack.

Try the sample on load: Halloween party, Witchy glam, $25-$50, Chilly outdoors, gives Midnight Witch in Velvet at $42.99, with a warm cloak layer for the cold.

How we built it

Plain HTML, CSS, and vanilla JavaScript, with no backend, no login, and no API keys. The costume library is deterministic, so every one of the 180 picker combinations returns a real plan, never a blank result. Automated tests check that every breakdown sums exactly, every total is inside its budget band, and every plan has a checklist.

How Kiro was used

Built spec-first with Kiro: a spec in /.kiro/specs/haunted-wardrobe/ with requirements, design, and tasks, steering docs in /.kiro/steering/ for product, tech, and structure, and a hook in /.kiro/hooks/ to run the tests when app.js changes. The spec kept the picker flow, cost maths, and checklist consistent, which is faster and safer than vibe coding without a plan. The /.kiro folder is at the root of the public GitHub repo and is not gitignored.

Challenges we ran into

The hard part was making every combination work and every price add up. A costume that fits a $25 budget cannot just be a $50 costume with a smaller number on it, and a chilly-outdoor plan needs a layer that a warm-indoor plan does not. We solved it by choosing a base plan by vibe and occasion, then scaling costs to the budget band and adjusting layers for comfort, with tests for all 180 combinations.

What we learned

A good costume plan is an explanation problem as much as a styling problem. People trust a plan when they can see why each piece is there, what it costs, and what to pack, not just a pretty picture.

What's next

Host the demo, publish the public GitHub repo, and record a short demo video showing the sample plan from picks to checklist.

Built With

Share this project:

Updates

Submission history