Inspiration
Every indie hacker ships a great feature, records a quick screen capture to show it off — and the result looks nothing like the product deserves. Shaky cursor, dead zoom, a raw window floating on a gray desktop. The polished demos you see from bigger teams take a video editor and hours in After Effects. I wanted the gap between "I just recorded it" and "this looks like a real product launch" to close to zero — on a Mac, in one app, for people who hate editing.
What it does
Screenify Studio is a Mac screen recorder that turns a rough capture into a polished product demo — and stages it inside a real iPhone or MacBook frame.
- Records the screen, then applies cinematic motion and smart auto-zoom that follows the cursor so viewers' eyes land where the action is.
- Drops the recording into device mockups (47 iOS frames + Mac frames) with a multi-device mockup studio.
- On-device features: auto-captions, background removal, and voiceover run locally on the Mac.
- An AI web-demo mode drives a real browser to record a flow (uses your own API key, not on-device).
- Ships an MCP server, so Claude, Claude Code, or Cursor can drive a recording programmatically.
- Free plan to start; Pro unlocks the higher-end export and polish.
How I built it
Native macOS app built for the Apple ecosystem — screen capture through the platform's own recording APIs, and a Metal-backed export pipeline so the motion, zoom, and device-frame compositing render fast at full quality. The web-demo mode automates a real browser and captures it as a first-class source alongside the screen. The MCP server exposes the recording engine over the Model Context Protocol so an AI agent can start, stop, and export a recording as structured JSON commands. The landing, blog, docs, and changelog run on Next.js.
Challenges I ran into
- Making auto-zoom feel intentional, not seasick — the cursor-follow motion had to ease and anticipate, not chase every twitch.
- Compositing a live recording into a device frame at export quality without dropping frames — this is where the Metal pipeline earned its keep.
- Deciding honestly what runs on-device vs. what needs an external API. Captions, background removal, and voiceover are local; the AI web demo calls out with a user-supplied key — and saying that plainly instead of over-claiming "all AI on your Mac."
- Exposing the recorder over MCP cleanly — every command speaks JSON, so an agent gets deterministic output instead of screen-scraping a UI.
Accomplishments that I'm proud of
- A recording goes from raw capture to a framed, zoomed, captioned demo in one app — no round-trip to a video editor.
- Being one of the first screen recorders an AI agent can actually drive, through MCP.
- Shipping continuously — the changelog is a real trail of releases, not a launch-day snapshot.
- A Mac-only focus that let the app lean on Metal and native capture instead of a lowest-common-denominator cross-platform stack.
What I learned
- For this audience, "less editing" beats "more features." The win is removing steps, not adding knobs.
- Be precise about AI claims — technical users spot "AI-powered everything" instantly and it costs trust. Naming what's on-device vs. API-backed is a feature, not a caveat.
- Distribution is a funnel problem, not a reach problem: getting people to the page is easier than earning the download.
What's next for Screenify Studio
- Deeper agent workflows over MCP — richer commands so an AI can assemble a full demo, not just a clip.
- More device frames and mockup scenes.
- Growing the on-device AI set where it can stay local.
- Screenify Mobile.
