What problem does it solve? Real desktop screen capture for terminal demos is flaky: window positioning races, focus issues, privacy leaks from visible apps, and non-reproducible output. This Skill lets you decide when to synthesize a terminal recording in Remotion instead of capturing a real screen, and shows how to author deterministic, narration-synced terminal scenes. ## Core Features & Use Cases - Decision heuristic: Clear criteria for choosing synthetic Remotion rendering versus real capture via screen_recorder, Windows-MCP, Cap, or Playwright. - TerminalScene component: A React component that renders a fake terminal with character-by-character typing, output lines, pauses, and floating pill badges, driven by a declarative steps array. - Narration pacing rules: A timing methodology (plus a trace script) ensuring every typed command lands exactly on its narration cue instead of burning through content early. - Use Case: You need a 60-second install walkthrough where each git clone and npm install command appears exactly when the voiceover mentions it. Author the steps array with pauses aligned to narration cues, render in Remotion, and get pixel-identical output on every re-render. ## Quick Start Ask the agent to create a synthetic terminal recording in Remotion using TerminalScene that types your install commands synced to the narration timestamps.
