What problem does it solve?
Turning newly recorded talking-head footage into a review-ready vertical video normally requires manual timeline editing, captioning, audio leveling, and format checks. This Skill automates the first editing pass by keeping creative decisions in a JSON edit plan and pixel operations in a deterministic FFmpeg renderer.
Core Features & Use Cases
- Explicit Edit Plans: Define segments, hook cards, captions, and audio targets in a validated JSON plan, with clean and aggressive variants for comparison.
- Deterministic Rendering: Trim and concatenate segments, rasterize captions and hook cards with Pillow, composite overlays, normalize loudness, and export 1080x1920 H.264/AAC MP4 via FFmpeg.
- Built-in QA: Automatically probe the export, verify codec and dimension checks, and capture three QA frames with a qa-report.json.
- Use Case: A founder records three takes of a new short-form script. Use this Skill to inspect the takes, build an edit plan from the transcript, and render a captioned, loudness-normalized vertical draft for Instagram Reels or YouTube Shorts review.
Quick Start
Use the net-new-video-editor skill to turn my new talking-head recordings in this project folder into a review-ready vertical video draft with captions and a hook card.
