Kinetic Editorial Skill — Development skill for Claude Code
Claude Code skill: word-by-word kinetic caption reels (9:16) built with HyperFrames + Three.js — storyboard, asset prep, offline SFX, build & preview.
How to install Kinetic Editorial Skill
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open ouerf-man/kinetic-editorial-skill and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Kinetic Editorial Skill does
Claude Code skill: word-by-word kinetic caption reels (9:16) built with HyperFrames + Three.js — storyboard, asset prep, offline SFX, build & preview.
Alternatives in Development
- Mono Color Skill — One-ink editorial print image skill — warm paper, halftone photography, active negative space, and restrained 3k ★
- Erduo Broll Loop Engineering — SRT 驱动的双后端 B-roll Agent Skill:自动路由 HyperFrames / Remotion,集成 152 张 Shotcraft 镜头卡 166 ★
- 07 Redesign Editorial — Redesign Inkwell as a typography-first editorial instrument — the feeling of drafting in a 143 ★
README
kinetic-editorial-reel
A [Claude Code](https://claude.com/claude-code) skill that turns a topic, script or voiceover into a vertical (9:16) **kinetic editorial reel**. Every spoken word appears on screen as it's said, with one red keyword per line and a literal hero object for each phrase. Scenes alternate between warm cream paper and black glow, with slam hits, camera pushes and sound effects. It's built with [HyperFrames](https://hyperframes.heygen.com) and Three.js.
What the skill does
- Intake: asks for the topic or message, the voiceover and transcript, the edit speed (calm / standard / rapid), the assets and the sound (SFX only, or with music).
- Storyboard: writes a frame-by-frame plan, with an image-generation prompt list if assets are missing. You approve it before anything is built.
- Preparation: subagents handle two jobs in parallel:
- Generate an offline sound-effect pack (
scripts/make_sfx.py). - Measure anchor points on your images so props land exactly where they should.
- Meanwhile the main session prepares assets, word timings and fonts:
- Assets: transparent PNGs are used as-is. Normal images get plain backgrounds removed
(
scripts/prep_assets.py). - Word timings: estimated from the voice's loudness envelope, with no speech-to-text
model (
scripts/word_timing.py).
- Assets: transparent PNGs are used as-is. Normal images get plain backgrounds removed
(
- Generate an offline sound-effect pack (
- Build: one HyperFrames composition with word-by-word captions, a GSAP timeline and
Three.js layers where 3D helps. Sound effects are placed automatically (
scripts/place_sfx.py). - Verify and preview: lint, snapshots,
hyperframes check, then the Studio preview. It renders only when you say so.
Requirements
- Claude Code
- The HyperFrames skills:
npx hyperframes skills update general-video - Node.js (for
npx hyperframes),ffmpeg, and Python 3 withnumpy,scipy,Pillow:pip install numpy scipy pillow
Install
Clone into a skills folder:
# for one project only
git
Related Skills
Reels Dopamina
Skill do Claude para criar Reels de alta retenção: roteiro, narração e animação HyperFrames
Hebrew Video Editor
Professional Hebrew (RTL) video editor skill for Claude Code — brief→script→storyboard→render, with Remotion +
Comment On Docx
Claude Code skill for adding editorial comments to Word (.docx) documents
Debug Asset Loading
Diagnose asset loading failures in Hatch Studios preview
Attack Tree
Build a formal attack tree for a named asset: maps all attack paths from root goal to leaf conditions, with AN
Xiaohongshu Prohibited Words
小红书违禁词离线检测(单一真相源):纯本地词库 + 扫描脚本,不依赖付费 API,覆盖免费课程/证书等场景的实测限流词。发布前必跑,退出码 0 才可发布。Offline prohibited-word checker f
Related Agents
Video Montaj
Video montaj va motion agenti. Use for any request to edit a talking-head / interview / Reels video — transcri
Motion Builder
Builds the HyperFrames composition — scenes, kinetic text, and data graphics — using the XChart library and th
Mf Editor
motion-forge social editor. Reads a word-timed transcript (words.json) and a goal (platform, length, tone) and