FluidEdits captures what you mean from the live preview: the target UI, your spoken instruction, and enough context for an AI editor to act.
From capture to pasted prompt in seconds — stay in flow while you iterate.
Press Alt+Shift+E and point at the UI element — a button, a banner, a table row. FluidEdits identifies the target and captures context automatically.
Tap the mic and say what you want: "Convert the first column to uppercase" or "Sort by quantity descending."
FluidEdits puts the instruction on your clipboard — paste into Cursor, VS Code, Lovable, or any tool that accepts text. Your change lands where you want it.
The extension turns each capture into a single, structured instruction — target, spoken intent, and enough context to act — and puts it on your clipboard. Paste into Cursor, VS Code, Lovable, or any chat or editor pane that accepts text — one portable format, no vendor lock-in.
Describe intent in plain language. No special syntax, no commands to memorize. Just say what the content should become.
Make multiple edits in one session. Send them all at once — your editor receives the full multi-region context as a single coherent instruction.
Edits queue up when your editor is busy. The extension badge shows pending count — nothing is lost, nothing fires out of order.
When your editor saves the change, FluidEdits can reload the preview tab you were pointing at — so the page catches up without an extra click. Same URL, fresh DOM, right after the patch lands.
Works on
Point FluidEdits at any UI component in your preview