Click the UI that needs work
Capture the real element, selector, bounds, styles, and source hints from the page in front of you.
Extension-first tooling for coding agents
Quey is the in-browser design-to-code layer for any production codebase. Build your imagination, fine-tune AI-generated interfaces visually and queue the payload to your agent.
Already loved by 100 people
Bring your own agents
Every surface
Select on the live page, stream context through MCP, and hand off to Cursor, Claude, or Codex — without leaving your workflow.
In your browser
In your repo
$ npx quey.devMCP · localhost:4747
In your agent
Same capture, your choice of harness.
Get started in seconds
Add Quey to Chrome from the Chrome Web Store and pin it for quick access. Chrome Web Store
Run the local provider in your project directory.
npx quey.devPress Alt+S to select any element and hand exact UI context to your agent.
Select, tune styles, and hand off to your agent
From screenshot to source
Capture the real element, selector, bounds, styles, and source hints from the page in front of you.
Keep the screenshot, DOM path, and component clues together so your agent starts from evidence.
Move from visual feedback to source edits without rewriting the request or hunting for files.
Review the final diff, commit, and deploy — knowing the change matches exactly what you saw in the browser.
Features
Point at any element on a live page, capture structure and styles in a shadow-root toolbar, and hand a clean payload to your editor or MCP bridge; selectors, bounds, and context stay accurate while you iterate.
Click any element on a live page to select it. Quey detects the React component, file path, and line number automatically.
Learn more →Drag a screenshot region and keep the same semantic context: primary element, related elements, stack hints, and agent-ready JSON travel with the image.
Learn more →Describe your changes visually and watch your coding agent edit the source file in real-time with streaming diffs.
Learn more →Workflow with less friction
Select live UI and capture the DOM path, bounds, and source hints.
Attach screenshot regions to the exact element and surrounding page state.
Send structured selectors, styles, and metadata to your agents.
Give Codex, Cursor, Claude, or any agent enough context to patch the right file.
Jump from the element to filenames, selectors, and component identity when source is visible.
Access frame, font, and color controls all in one place while staying on the selected element.
Pricing
Start with the core browser workflow, unlock Pro for active usage, or buy Lifetime once.
3-day free trial, then $5/mo. Cancel anytime.
What's included
$79 launch pricing
One payment for the full editing workflow.
What's included
7-day full refund guarantee
Everything you need to know
Yes. Quey is the visual handoff layer. Bring your preferred coding agent, then use Quey to capture the live element, DOM path, and source hints before you ask for the edit.
Yes. You can start with the core browser workflow for free, then upgrade later if you need more active usage or a lifetime unlock.
No. The same capture workflow can be used with Claude, Codex, Cursor, ChatGPT, or any MCP-aware setup that can consume structured UI context.
Install the Chrome extension for selection in the browser, then run npx quey.dev inside the repo you want to edit so the provider can expose project-aware context to your agent.