Skip to documentation

Layrr

Layrr is for the moment when you can see the UI problem in the browser, but explaining it through filenames and line numbers is slower than pointing at it.

Run Layrr beside your local dev server, open the proxied app, select the part of the interface you want to change, and send that context to Claude Code, Codex CLI, or Gemini via Pi. The project stays in your repository; Layrr is the browser-side editing layer.

I need to change one visible element

Use visual editing when the problem is spatial:

Select the element in the browser, drag or resize it, preview the change, then apply it so the codebase is updated.

I need to fix copy without hunting through files

Use text editing when the problem is content:

Click the text in the browser, edit it inline, and apply the update. This is most useful when the same words may be buried inside components, route files, or data objects.

I need to change a whole section

Use area selection when the problem is not tied to one DOM element:

Draw a rectangle around the region, describe the outcome, and let the coding agent inspect the selected UI in context.

I need to match a mockup

Use design-to-code when you have a target image:

Upload the image with a short instruction. Layrr gives the agent both the visual target and the current app context, so it can update the existing code instead of generating an isolated snippet.

How Layrr fits into your dev loop

Layrr sits between your browser and your app:

By default, the proxied app opens at http://localhost:4567.

What ships in the repo

Next steps