What problem does it solve? Turning a vague request like "make me a video" into a finished, rendered video requires choosing the right pipeline, gathering the right inputs, and coordinating many moving parts. This Skill acts as the mandatory entry point for any video, animation, or motion-graphic request, inspecting project state, capturing intent through a structured interview, and routing to the correct HyperFrames workflow. ## Core Features & Use Cases - State-aware routing: Detects whether a request is a fresh creation, an edit, a project operation (validate, preview, render, publish), or a Remotion port, and resumes from BRIEF.md, hyperframes.json, or STORYBOARD.md when they exist. - Intent interview and pitch round: Converts unformed requests into a confirmed BRIEF.md through memory lookups, recipe adoption, five-concept pitch sampling, must-have questions, and run-shape choices (storyboard, automation vs companion). - Workflow and domain skill orchestration: Routes to ten workflows (product-launch-video, faceless-explainer, pr-to-video, music-to-video, motion-graphics, slideshow, embedded-captions, talking-head-recut, remotion-to-hyperframes, general-video) and lazy-installs them plus domain skills via npx hyperframes skills update. - Use Case: A user says "turn this GitHub PR into a changelog video." The Skill routes to /pr-to-video, reads the route contract, asks the must-have questions with a length recommendation derived from the PR's diff size, writes BRIEF.md, installs the workflow, and hands off execution. ## Quick Start Ask the agent to create a video from your website URL, GitHub PR, music track, or existing footage and it will route, interview, and build it with HyperFrames.