kapishdima

kapishdima / whichly

Public

Code multiple variants of a section, let clients pick the winner live on your real staging site

www.whichly.xyz
11
0
89% credibility
Found May 29, 2026 at 11 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

Whichly is an open-source tool for React websites that lets developers show multiple versions of any design section side-by-side. Instead of going back and forth in Figma or guessing what a client wants, you write two or three versions directly in your code, deploy to staging, and let the client flip between them live on the actual page. When they pick their favorite, the exact selection is saved in the URL so it can be shared easily. The picker appears as a floating toolbar and works in Shadow DOM so it never breaks your site's styles. No backend, no dashboard, no account needed—just one React package.

How It Works

1
🔍 Discovers Whichly online

A developer stumbles upon Whichly while searching for a better way to get client feedback on design choices.

2
📦 Adds the tool to their project

The developer installs one small package and wraps their app with a simple provider component.

3
🎨 Creates multiple versions of their design

For any section they want to test, they wrap it in a Block and add two or three Variants with different looks.

4
🚀 Deploys to a staging site

They push their code to a staging environment just like they always do—no special setup needed.

5
Shares a link with the client
💻
Client opens on desktop

Client views the page on their computer and clicks through variants using a floating toolbar

📱
Client opens on mobile

Client views the page on their phone and uses the same floating toolbar to pick their favorite

6
👆 Client clicks through and picks a favorite

The client sees each variant live on the actual page, in the real layout, and clicks to choose the one they like.

✅ Everyone agrees on the winning design

The client shares the link showing their choice, the developer removes the rejected variants, and they ship with confidence.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 11 to 11 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is whichly?

Whichly is a React library that lets you code multiple versions of a UI section, then let clients toggle between them live on your actual staging site. You wrap blocks in `` and `` components, deploy to staging, and share a link. The client clicks through options until they land on the one they want. Variant state lives in the URL, so any selection is just a shareable link. No backend, no dashboard, no auth. The picker mounts in a Shadow DOM, so its styles never bleed into or out of your page.

Why is it gaining traction?

This solves a specific, painful problem: the Figma-to-production handoff where clients approve mockups that look nothing like the final site. Instead of arguing over static images, they see real components in real layouts with real data. The URL-encoded state means you can A/B test with a single client without any analytics infrastructure. It's a clever pattern that replaces an entire workflow tool with a 200-line React component.

Who should use this?

Frontend developers building client-facing sites who are tired of screenshot reviews and endless revision rounds. Freelancers and small agencies who want clients to self-serve design decisions on staging. Anyone doing design-by-committee where stakeholders need to see actual implementation, not mockups.

Verdict

Whichly is a clever, focused tool that does exactly one thing well. The API is clean, the Shadow DOM isolation is thoughtful, and the URL-based state is elegant. At 11 stars, it's early and unproven at scale. The docs exist and the code looks solid, but there's no visible test suite and the project is single-maintainer. The 0.9% credibility score reflects that maturity gap. Worth trying on your next staging review, but don't bet a production launch on it yet.

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.