What problem does it solve? Manually designing animated overlays for talking-head videos means picking effects, timing them to speech, and re-rendering on every change. This Skill reads an SRT subtitle file and produces a ready-to-import overlay JSON where card selection, timing, and pacing are derived from the actual transcript. ## Core Features & Use Cases - SRT-to-overlay orchestration: Parses subtitle timestamps, segments the script into chapters and semantic sections, then assigns cards from a fixed library of 20 effect types (chapter bars, stat counters, quote cards, timelines, bilingual captions). - Density, diversity, and layout rules: Enforces pacing constraints such as no adjacent identical card kinds, per-segment accent colors, and rhythm parameters computed from real SRT durations. - Lint self-check before delivery: Runs the overlay linter to catch placeholder text and out-of-range timestamps, and preserves user-set lintOff overrides. - Use Case: Export subtitles from a talking-head video, hand the SRT to your agent, and receive a video-overlay.json you import into Overlay Studio to preview, tweak, and export as a transparent MOV layer. ## Quick Start Give the agent your SRT subtitle file and ask it to generate an overlay JSON for Overlay Studio using this skill.
