What problem does it solve? Turning a proven reference video into a new original video usually means manually reverse-engineering its structure, planning shots, generating assets, and assembling a timeline. This Skill orchestrates that entire clone workflow inside the Recut editor: it reads the reference, re-directs a new film around your subject, product, language, or CTA, and delivers an editable, exportable project. ## Core Features & Use Cases - Reference-to-Treatment Pipeline: Imports and transcribes the reference video, records its on-screen text system and camera language in project markdown files (BRIEF/reference/TREATMENT/PLAN/PROGRESS), then derives keep/replace decisions for the new film. - Scene-Level Generation: Generates assets in scene/segment units rather than isolated clips, using storyboard references, style/motion anchors, first-last-frame contracts, and World voice references for character dialogue. - Anchor-Based Timeline Assembly: Maps speech@ and clock@ anchors to timeline windows via the timeline-editor (recut.editor.*), validates with timeline.validate, and exports only after actually watching the result. - Use Case: You have a competitor's viral product video and want the same structure with your own brand and voice. The Skill builds the project, writes the treatment and shot plan, gets your budget approval, generates each scene, assembles the timeline, and exports the finished cut. ## Quick Start Clone this reference video into a new film for my product, keeping its structure and camera language but replacing the subject, script, and on-screen text.