Click it. AI gets it.
Select any element to automatically capture its source code, component, state, and relevant context. No need to tell AI where the file is or which component to look for—just start the conversation.
No digging through files, no screenshots, no explaining which button you mean.
2 steps: ① Install the browser extension ② Paste the prompt into Claude Code / Cursor
Manage your plan and payment method
Select an element on the page and Curxor automatically picks up the related source and component context. No digging through files, no screenshots, no explaining “that button in the top-right corner” — just tell the AI what you want changed.
Select multiple elements at once and tell the AI everything you want changed. Cross-component tweaks, batch style updates, page refactors — no more doing it one element at a time.
Select an element and the page visual goes to the AI together with its code context. Just say “this spacing is off” — the AI already knows exactly where you mean.
No manual screenshots, no image uploads, no struggling to describe where the problem is.
Frame any region of the page and mark the problem with arrows, rectangles, or a pen. The screenshot and your annotations go to the AI together, so it sees exactly what to fix.
Circle it and send — no need to type out what you marked.
Freely switch between parent, child, and nearby elements to keep the context just the right size. Less irrelevant code means sharper understanding and fewer wasted tokens.
All source analysis happens locally — your project code is never uploaded. Once the local Agent is connected, Curxor can jump from the running page straight to your source.
See the file and code location behind any element, then jump straight to that file and line in Cursor or VS Code. No searching, no guessing where the component lives.
Browser, source, and editor are one step apart — save your change and the page updates instantly.
Select once — source, context and screenshots arrive together. Fewer round-trips, less wasted time and tokens.
Select any element to automatically capture its source code, component, state, and relevant context. No need to tell AI where the file is or which component to look for—just start the conversation.
Select multiple elements and describe what needs to change. Send everything to AI in one go for multi-component edits, refactors, and styling updates.
Capture an element screenshot with one click, or frame any area of the page — send the visual together with its code context. No manual screenshots, no uploads, no struggling to explain what's wrong.
Source analysis and context extraction happen entirely on your machine. Your code doesn't need to be uploaded to a server—keeping development fast, private, and secure.
Auto-detects the page framework and component stack — no extra setup. Spend your time solving problems, not configuring frameworks.
Record your actions and give AI the full context of what happened. Reproduce bugs and explain multi-step interactions without having to describe every step yourself.
Install the browser extension, connect the agent you already use, and you're ready to go.
2 steps: ① Install the browser extension ② Paste the prompt into Claude Code / Cursor
Free to use · ready right after install