simon-skills
Skills for faceless video channels: one topic becomes a video with voiceover, subtitles, covers and per-platform copy.
- 520 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- Runs Node, FFmpeg and Playwright, writes project files and calls speech synthesis services with your keys.
- Videos naming brands need fact checking: the author is responsible for the claims.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/simon-skillsDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- Runs Node, FFmpeg and Playwright, writes project files and calls speech synthesis services with your keys.
- Videos naming brands need fact checking: the author is responsible for the claims.
This is third-party code. Review the repository files before installing.
What it does
simon-skills is a set of three skills for Claude Code and Codex that produce videos for faceless channels. `investigation-video` builds an investigation of ten minutes or more: a fact chain, real footage, Remotion editing, music ducked under the voice, copy for eight platforms. `whiteboard-video` draws a two-minute whiteboard explainer with stroke-by-stroke animation, stickers and subtitles. `video-common` keeps the shared steps: fact checking, a source ledger, cover review and delivery review. The agent follows `SKILL.md` step by step and can pause at any of them for edits. Work ends with a finished video and a materials package; publishing is left to a human.
Who it is for. Video channel authors who publish Chinese-language videos without a camera or manual editing.
Good fit when
- A market or consumer topic becomes a ten-minute-plus video with source references.
- An AI news item is explained in a two-minute whiteboard video.
- A finished video needs 4:3 and 3:4 covers and platform copy.
- Each claim is checked against the source ledger before release.
Not a fit when
- The video must be in Russian: voiceover, templates and copy target Chinese and Chinese platforms.
- The machine has no Node, FFmpeg or Playwright and building that environment is not an option.
- Releases must publish themselves: the skills stop at the finished package.
Example request
Make a whiteboard video on why plans do not survive the week. I need voiceover, subtitles and covers.Limitations
Voiceover needs Volcano Engine keys in each skill's `.env`. Building requires Node, Remotion 4.0.508 with React 19.2.8 (the author advises against upgrading), FFmpeg and Python, and the whiteboard skill also needs Playwright and Codex CLI. Tested on macOS; web chat will not work. The repository ships no music: bring your own royalty-free track. News footage is acceptable only in a review context with a source credit, and the video author carries the rights responsibility.
How to disable. Remove the `investigation-video`, `whiteboard-video` and `video-common` links from `~/.claude/skills` or `~/.codex/skills`. Delete the `.env` files holding speech synthesis keys separately.
FAQ
Do I need to show my face or edit video?
No. Frames are assembled from materials or drawn in code, and the voice is synthesised. You can clone your own voice in the Volcano Engine console.
Does the skill publish videos itself?
No. Work ends with a finished video and a publishing package; the author decides what happens next.
Why must Remotion stay on one version?
Upgrading caused animation timing glitches, layout drift and black frames. Version 4.0.508 with React 19.2.8 removed them.
How does the skill treat footage rights?
News footage is used only in a review context and labelled with source and date. The rules are in `skills/video-common/references/compliance.md`.
What does voiceover need?
Volcano Engine keys. They go into a `.env` file inside each skill, following `.env.example`.
Discussion
No questions or reviews yet.
Related
SkillsA set of SKILL.md skills for short-video and content production: topic, script, voice, avatar, editing, covers and motion graphics
A Claude Code and Codex skill that turns a topic into a narrated motion-graphics explainer video on Remotion with subtitles and chapters
The video-explainer skill turns a script into a vertical 9:16 explainer: narration, captions, Remotion and a pre-render review
A skill that turns a topic into a narrated 4K video: research, script, local TTS, Remotion composition and an FFmpeg mix
A skill for Claude Code and Codex that builds cinematic product videos with Remotion from shot recipe cards, a ready template and audio assets
A video editing skill for Claude Code and Codex: the agent cuts pauses and filler words, grades color, burns subtitles and adds animated overlays
README in short
The README presents the set as skills from the working SimonTalk channel: the author says more than 80 videos were released with them as of October 2026. Each skill gets a path from topic to finished release and a stage table. Separate sections cover platforms, FAQ and material rights.
README badge
Original SKILL.md
--- name: investigation-video description: 制作或精修不露脸商业与消费调查视频，覆盖选题、事实核查、10分钟以上叙事、真实动态素材、配音字幕、CTA，以及AI封面和各平台发布文案。按当前请求执行阶段，保留已确认稿件和品牌；不自动公开发布。 --- # 调查叙事视频 目标：故事由具体画面推进，机制由 MG 解释，结论有出处，品牌与账号配置保持一致。本包的 **30 秒视觉升级 v2 样片**保留为拼贴质感参考；长片同时遵循《零食称重》制作中确认的真实素材、动态叙事与发布要求。样片不代表长片素材已经够用。先看相应参考成品，再设计；参考的认可范围见各文件，不把一次认可扩大到全部后续版本。 ## 先确定本轮范围 - 用户要求学习对标的“画面／质感”时，围绕构图、素材合成、景深、边缘、字体、光影与动态制作展开；不要把交付转成选题、文案或叙事结构优化。 - 延续已有选题、旁白、事实口径、画幅和已确认版本。视觉精修不等于重写稿件或更换音色；用户明确指出“没有钩子／重做开头”时，可在开头范围内改稿、重配音并检查正文衔接，保留原声线。 - 新做完整长片时，以用户时长为准；本流程的账号目标为 **10 分钟以上**，可规划约 10–12 分钟，**且按账号默认 1.2 倍语速（火山 speech_rate=20）实测**，见[素材、事实与声音](references/editorial-and-materials.md)。样片通常 20–30 秒；不拿样片时长当长片要求，也不靠重复画面或放慢语音凑时长。 - 本次认可样式为横屏 16:9。其他画幅按用户和当前工程适配，不因为旧技能默认 dock 而自动增加上下品牌带、章节和进度条。 - 用户已认可方向就继续制作。只有更换主要视觉语言时，优先做一段动态样片验证；一般视觉样片覆盖实拍、证据与MG，开场样片按钩子需要编排，不强塞三种形式；不要每次常规修改都重新索要审批。 ## 按阶段推进 新选题、完整长片或续作先读 [长片制作流程](references/long-form-workflow.md) 与 [管线手册](references/pipeline-playbook.md)。恢复当前选题、确认稿、音轨、素材、成片和发布包的版本；“继续”接着未完成项做，不重新选题或重做已确认内容。 - 只要求选题：交付有事实入口、素材可得性和传播钩子的候选，每个候选标注核心问题类型（见[叙事模式](references/narrative-modes.md)）和当下热点关联，热点题排前；不自动制作整片。 - 完整制作：选题与事实 → 长稿与钩子 → 素材覆盖 → 配音字幕与剪辑 → CTA → 检查与交付；已有确认环节直接复用。 - 只改视觉、CTA或发布材料：只做对应部分。发布包需求读 [AI封面与发布文案](references/covers-and-publishing.md)；“完整交付”默认包含成片、字幕、来源与检查说明、两种AI封面和适配文案。 - 不用普通确认点中断已经授权的制作；缺少影响事实、范围或执行的必要信息才提问。 ## 项目结构 一期一个工程目录（`projects/<date>-<slug>/`：work/research、work/production、work/covers、outputs）加一个知识库目录（`knowledge/<date> <题名>/材料/`）。正文、来源、文案、封面、检查只在知识库维护，工程 outputs 用软链接；改镜头表后清空裁切目录重切；旧版本不删。完整树与版本规则见 [项目文件夹结构](references/project-layout.md)。 ## 开场先建立观看理由