Chapter 05 · Our Favourite Skills · Reading 05

The HyperFrames Student Kit, Nate Herk's editing pipeline

Radiant Digital · ~2 min read

The 14 skills that turn raw talking-head footage into a finished reel, cut, captioned and story-planned. By Nate Herk, MIT licensed. Our short-form scripts follow his structure.

What it is

HyperFrames renders video; this kit teaches an agent to edit one. Nate Herk built it for his students: bring your own footage, and the skills transcribe it, cut the dead air, find the flubbed takes, plan the story, and build motion graphics on top with HyperFrames and GSAP. It also carries 406 draft motion-graphics cards and two scene templates you can reuse.

Our own short-form process is built on it. Every script we write follows the kit's rules: the topic is clear by second three, there is one honest open loop with a visible object, the payoff is earned, and there is exactly one call to action.

Author: Nate Herk. MIT licensed.

Repository: github.com/nateherkai/hyperframes-student-kit

The skills inside it

• cut-silences. Agent 1 of the editing pipeline. Removes silences and dead air from a recording using a word-level transcript, and writes an edit list the next agent can read.
• cut-mistakes. Agent 2. Finds stutters, repeated words, false starts and re-recorded lines, proposes every cut with context for your approval, then renders only the approved ones.
• edit-video. The full edit, start to finish: transcription, silence trimming, mistake review, visual storytelling, motion graphics, verified render.
• short-form-edit. Reels, Shorts and short ads: curiosity-led openings, earned payoffs, story-driven cuts, transcript-synced graphics, moving B-roll, captions, music and sound effects.
• video-storytelling. Treats the visual layer of a long talking-head video as one persistent world the camera moves through, instead of a stack of separate graphics. The fix for edits that feel "clearly made by an AI".
• hyperframes-video-beats. Plans, times and validates the motion-graphic beats in a long video: lower thirds, callouts, full-screen takeovers, retention graphics.
• style-library. Chooses, customises and extends the reusable card library and style templates.
• gsap. A focused GSAP reference: tweens, easing, stagger, timelines, performance.
• make-a-video. The beginner's door: one interview, then it builds the whole video with preview and visual-check gates.
• website-to-hyperframes. Captures a website and makes a video from it.
• Also: hyperframes, hyperframes-cli and hyperframes-registry (the kit's own copies of the core skills) and short-form-video, which maintains the kit's original teaching compositions.

Install it

git clone https://github.com/nateherkai/hyperframes-student-kit.git
cd hyperframes-student-kit
npm ci
npm run setup
npm test

The kit's transcription helper uses ElevenLabs by default; you can point it at OpenAI Whisper or a local Whisper instead, which is what we do.