SkillFoxx

Motion Skill

A Codex skill that turns a script or a voice recording into video with motion graphics, captions and an editable Remotion project.

  • 127 ★ on the repository
  • Last commit
Skill

Extended access

Why this level

  • The skill downloads stock media from the web and licenses must be checked by you.
  • Installing dependencies runs a Node.js build and a large music model download.
All reasons and checks

derek-zhuolin/motion-skillAuthor:derek-zhuolin

Install

In your terminal, with SkillFoxx CLI

npx skillfoxx add skills/motion-skill

Detects the agents on your machine, shows the access level and pins the version.

Other ways to install

Access and checks

How we review

Extended accessThe tool runs code or makes network calls and only reads external data.

  • The skill downloads stock media from the web and licenses must be checked by you.
  • Installing dependencies runs a Node.js build and a large music model download.

This is third-party code. Review the repository files before installing.

What it does

Motion Skill builds a video without the author on camera. The user provides a script or a ready voice recording, the agent analyzes the meaning, lays out scenes and assembles video from real footage, graphics and titles. The result is a Remotion project that stays editable: scenes, timing and captions are files next to it. The repository holds 96 animation technique studies in 19 categories, 10 ready components and 5 kinds of semantic structure. Python checks timing and constraints, and Remotion renders the MP4. Music comes from local generation through a separately installed ACE-Step or from your own licensed track.

Who it is for. Authors and experts who want to publish videos from a script or voice recording without appearing on camera.

Good fit when

  • You have a podcast or voice note and need a video with titles from it.
  • An idea has to be shown through diagrams and animation without a talking head.
  • A finished MP4 is needed together with an editable project and material sources.
  • Synchronized captions at the bottom of the frame are needed over a voice track.

Not a fit when

  • You need live-action editing with cinema-grade effects.
  • Node.js, FFmpeg and a Chrome browser are missing on the machine.
  • You need license-clean media without checking sources.

Example request

/motion-video Make a video from this recording without my face: keep the voice, add captions and a closing title.

Limitations

The version is a public beta and the license is marked NOASSERTION, so read the terms in `LICENSE_SCOPE.md`. Node.js 22 or newer, Python 3, FFmpeg and Chrome or Chromium are required. The ACE-Step music model installs separately and takes about 10 GB of disk plus cache, and the first download is slow. The agent finds stock media online, so verify licenses and authors before publishing. The timing example uses a test rhythm and does not guarantee the quality of your video.

How to disable. Remove the skill with `npx skills remove` or delete the skill folder by hand. The Remotion project with video lives separately and can be kept or deleted.

FAQ

Which command installs the skill?

`npx skills add derek-zhuolin/motion-skill --skill motion-video -g`, then choose your agent. Node.js 22.20 or newer is required.

Is my own voice required?

No, a script or a ready recording works. If music is needed, ACE-Step generates it or you use a licensed track.

What is the output?

An MP4, an editable Remotion project, captions and notes on media sources.

How do I update the skill?

Run `npx skills update motion-video -g`. It replaces files, so back up your own edits first.

Does it work with Claude Code?

The README names it as a tool for Codex and agents. The Skills CLI can install it into several agents, but validation is stated mainly for Codex.

Discussion

No questions or reviews yet.

Skills

A skill that turns a coding agent into a studio for voiceover explainer videos: word-level sync, a motion library and Remotion assembly

SkillExtended access2.3K

A motion design skill: the agent writes Remotion video code with voice, music and sound and delivers only after measuring frames

PluginExtended access75

A skill for Claude Code and Codex that builds cinematic product videos with Remotion from shot recipe cards, a ready template and audio assets

SkillExtended access18.9K

A Claude Code and Codex skill that turns a topic into a narrated motion-graphics explainer video on Remotion with subtitles and chapters

SkillExtended access3.3K

A Remotion video starter: ten styles and two skills let an agent make a motion video from a repository link.

SkillExtended access54

README in short

The README covers script and voice-recording modes, 96 animation technique studies and the first-install order. A separate section explains how to install ACE-Step for music and how much space it needs.

README badge

Original SKILL.md

---
name: motion-video
description: 显式调用 Motion Skill / Motion video 时，优先继承已选影片的实际镜头与素材，制作连续品牌短片或原声解释视频；按需检查，不自动调用模型、上传或发布。
---

# Motion Skill · Motion video

公开测试版 **1.2.0-beta.1**。增加可选重点运镜、编排证据与音乐驱动律动；旧公开示范片、10个稳定 adapter 和既有工程保留。

## 默认入口：规范 → 编排清单 → 实现

只要本次要写或改 Remotion，先运行 `python3 scripts/motion.py preflight-remotion --project /absolute/project`，读取它定位的 **remotion-best-practices/SKILL.md**，再按任务读取具体分支。无需用户重复点名。只是找到文件不等于已读；记录实际读取路径、版本与 hash。缺失时按 `references/remotion-integration.md` 处理，不假称已调用、不静默联网安装。保留已可运行工程与内容准备工作。

动手前按 `references/semantic-orchestration.md` 给用户一张**简短镜头编排清单**：时间/本段意思、主焦点、主效果编号与名字、辅助效果、进出场和交接对象、选择原因、可读窗口。先说再做；这份清单不是新增逐镜审批。数字快照、数字变化、关键词层级、证据截图、主体运动的视频要分别判断。每镜一个主表达；效果数量随内容，不规定10或15组。

由当前 Agent 读内容和核对素材。对已覆盖的数字、关键词、真实素材等关系，填写 `assets/semantic-plan.example.json` 的结构，再运行 `select-motion plan.json --out selection.json --explain shot-plan.md`。这一步按语义与近期使用记录给适配候选，不能把同一套 AME 镜头顺序套到所有内容；相关性优先，变化不靠随机或硬凑。返回 planned 不代表已调用。实际导入 `production-registry.json` 中的组件后，记录 actual-used 清单并审查源代码与画面；没实现的研究保持 study_only。仅修改某一镜时只重选受影响镜头。新关系不必硬塞进这几类：可按 catalog 选研究并定制实现，在同一编排表中标 custom、来源编号和代码位置；用相关源码与实际画面验证，不冒称选择器或生产接线审计已通过。

## 先走最短的正确路径

读取项目 AGENTS、当前任务及用户选定的视觉基准。新 run 保留原件。本机已登记授权基准时运行 `python3 scripts/motion.py baselines`，有已选原片就使用 `reuse` 继承实际影片代码、字体和授权素材；**不要先用五类默认卡片模板重建它**。

- **连续 Launch**：沿用用户选定影片的视觉语言。让一个载体贯穿发现、选中、展开、协作；按语义选少量转折，保留阅读停顿。
- **原声解释**：沿用用户选定影片的字体、材质与节奏。同一视觉跨多句话，真实影像与示意关系分工；原声保持 1×，无视频不生成说话人脸。
- **修改/复盘**：直接读目标工程的相关镜头和时间表，不重跑转写、信源研究、模板生成或全研究库。
- **没有可用基准**：才用 `new` 建标注清楚的结构草稿，并按实际内容写镜头。默认五类 renderer 不是电影感模板，也不会自动调用研究适配器。

`reuse` 只复制本机已登记源，不执行源脚本，不联网、不转写、不渲染。新工程初始是 reference_copy，旧时间和旧口播属于旧片；换内容时更新语义事件、字幕和资产映射后才能称为新片。没有登记源时准确报缺失，不静默降级为卡片片。

```sh
python3 scripts/motion.py baselines
python3 scripts/motion.py reuse --baseline my-reference --dest /absolute/new-run --deps /absolute/existing-node_modules
# 基准需登记自己的授权工程；公开包不含作者私人基准。见 references/local-fastpatch.md。
```

## 语音：同一段只转写一次

先查已核对的项目 transcript 和 source hash；没有就查询缓存：

```sh
python3 scripts/transcript_cache.py get --media /absolute/audio
python3 scripts/transcript_cache.py put --media /absolute/audio --transcript /absolute/transcript.json  --engine existing-local-engine
```

缓存默认位于本 Skill 的 `.local/transcript-cache`，不进入分享包。命中只证明源与转写文件匹配，仍需检查语言、词时标、覆盖范围和专名；不自动把 verified 改为 true。只做短样片时优先处理选段及少量上下文，不默认整段转写两遍。复用现有可用本地引擎；启动前记录输入 hash、范围、引擎、预计时长和超时。已有转写任务仍在运行时不要并行开第二种实现；确需切换时先确认旧任务已退出，再只补缺失范围。不因尾部卡住重复跑完全部内容。本 Skill 不内置 ASR 模型或自动切换服务。

Report a problem