What problem does it solve? Turning a raw talking-head, interview, or podcast recording into a polished, watchable social video normally requires manual motion-graphics work in an NLE editor. This Skill automates the packaging step: it transcribes the video locally, plans timed graphic overlay cards (titles, lower-thirds, data callouts, quotes, side panels, picture-in-picture) synced to what is being said, and renders the result to MP4. ## Core Features & Use Cases - Transcript-driven card design: Extracts audio with ffmpeg, transcribes locally with Whisper via hyperframes transcribe, and derives card timing and content directly from the word-level transcript. - Design system with 10 styles, 4 layouts, 3 frames: A reference library (academic, editorial, minimal, spotlight, geom, whiteboard, audit, terminal, swiss, xhs) combined with split/stack/pip/overlay layouts and clean/hairline/polaroid video frames, plus GSAP-driven animation via data-anim attributes. - Density-aware pacing: Computes card count from video duration and information density, with a minimum of 5 cards and no upper clamp for long-form content. - Use Case: You have a 2-minute founder interview clip. The Skill asks you to confirm ratio (16:9 / 9:16 / 4:5), layout, style group, and card count, then writes each card's HTML, assembles a GSAP composition, and renders a packaged MP4 with lower-thirds and data callouts synced to the speech. ## Quick Start Ask the agent to add designed graphic overlay cards to your video file, for example: "Package my interview.mp4 with graphic overlays — titles, lower-thirds, and data callouts synced to what's being said."
