FluidEdits

Point an arrow at a UI element, speak your change, and copy a Gemini-built IDE prompt to the clipboard.

FLUIDEDITS — WHAT IT DOES FluidEdits is a browser extension for developers who already have a web app open and want to move faster from "I see the problem on screen" to "here is a precise, pastable instruction for my editor or AI assistant." Instead of typing long explanations about which button, card, or row you mean, you point with an on-page arrow, speak your change out loud, and copy a structured IDE prompt to the clipboard. An AI model (Google Gemini) uses your screenshot crop, audio instruction, and page-aware context so the prompt matches what is actually rendered in the browser. WHY INSTALL IT • Save time — Skip hand-writing selectors, layout explanations, and nested component descriptions. • Cut ambiguity — The arrow ties your request to a specific element in live DOM context, not a vague "the one on the left." • Better handoffs to AI tools — Output is meant to paste into Cursor, VS Code, or similar workflows so the next step is implementation, not back-and-forth clarification. • Sensible security model — The Gemini API key is not shipped inside the extension. Your generations go through FluidEdits infrastructure (Supabase Edge Functions) over HTTPS, so secrets stay server-side. HOW IT WORKS 1) Open the page you are working on (local dev server, staging URL, or most http(s) sites you use for UI work). 2) Press the FluidEdits keyboard shortcut to open the overlay on that tab. 3) Draw an arrow from your starting point to the exact UI element you want to change. 4) Record a short voice note: what should change (copy, spacing, visual bug, behavior, accessibility, etc.). 5) Send — FluidEdits captures the relevant screenshot region and audio, sends them to the FluidEdits backend, which calls Gemini and assembles a structured, paste-ready prompt. 6) Paste — The finished prompt is written to your system clipboard so you can drop it into your IDE, assistant chat, or ticket. TYPICAL USE CASES • Tweaking UI copy or spacing on a live preview without writing a novel in chat. • Capturing a concrete "change this element" request tied to what you actually see. • Turning an in-browser review into an actionable instruction for you or your team. • Keeping feedback grounded: less "it looks wrong" and more "this target, this intent." WHO IT IS FOR • Frontend and full-stack developers shipping web UIs. • Teams that already use AI coding assistants and want requests grounded in what is on screen. • Builders who prefer show-don't-tell when describing UI work. WHAT MAKES IT DIFFERENT FROM ONLY SHARING A SCREENSHOT A raw screenshot leaves your AI or teammate to guess which control you mean and what you want done. FluidEdits pairs a precise arrow-to-element gesture with your spoken instruction and structured page context, so the exported prompt is closer to a ready task than a mystery image. LICENSING AND PLANS Depending on how FluidEdits is configured for your installation, optional paid licensing (via Lemon Squeezy) may apply, including usage limits that reset on a defined schedule. The extension popup includes license activation and quota status where your publisher enables it. Internal or developer builds may skip license gating—follow your publisher's setup guide. WHAT YOU NEED • An active internet connection when sending a capture (the AI step runs remotely). • Chrome (Manifest V3). Microphone permission when you choose to record audio. • A FluidEdits backend configured by the publisher (the extension communicates with Supabase-hosted endpoints; you do not paste API keys into the extension). PRIVACY (SUMMARY) FluidEdits processes what you explicitly capture during a session—screenshot material from the visible tab, your voice instruction, and related page metadata useful for locating and describing the target element—only when you run the workflow. Data is transmitted over HTTPS. Processors may include Supabase (backend), Google (Gemini), and Lemon Squeezy (license commerce and validation) when paid plans are used. The Chrome listing links to a privacy policy with full detail on categories of data, retention, and your choices. PERMISSIONS IN PLAIN LANGUAGE The extension uses standard capabilities to make the workflow possible: access tied to an active, user-initiated session for overlay and capture, optional microphone input while recording, clipboard write to deliver the generated prompt, local storage for settings, license, and optional job queue data as implemented, and HTTPS calls to FluidEdits endpoints. Capture does not start until you use the shortcut and complete the on-page flow. HONEST LIMITATIONS • Quality depends on how clearly you describe the change and how readable the target region is in the capture. • The extension is built for developer workflows; it is not a general screen recorder or meeting tool. • Features such as licensing, quotas, and backend URL depend on how your publisher ships FluidEdits. GET STARTED Install FluidEdits, open your app in Chrome, use the shortcut, draw the arrow, record your note, send, and paste the generated prompt wherever you build or review code. For configuration, licensing, or support, use the contact information on this listing or your organization's FluidEdits documentation.