Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add chenyuxiaojin/cyxj-remotion-starter --skill cyxj-remotion-producegit clone --depth 1 https://github.com/chenyuxiaojin/cyxj-remotion-starterWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/chenyuxiaojin/cyxj-remotion-starter/cyxj-remotion-produce)<a href="https://agentmods.dev/skills/chenyuxiaojin/cyxj-remotion-starter/cyxj-remotion-produce"><img src="https://agentmods.dev/badge/skills/chenyuxiaojin/cyxj-remotion-starter/cyxj-remotion-produce/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/chenyuxiaojin/cyxj-remotion-starter/cyxj-remotion-produce"><img src="https://agentmods.dev/badge/skills/chenyuxiaojin/cyxj-remotion-starter/cyxj-remotion-produce.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00151 | $0.01384 |
| Opus 5 | $0.00076 | $0.00692 |
| Sonnet 5 | $0.00030 | $0.00277 |
| Haiku 4.5 | $0.00015 | $0.00138 |
Grade A, and why
cyxj-remotion-produce scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 44 lines — stays where its author put it; the contents beside it link to each section on GitHub.
cyxj-remotion-produce — 出片流程编排器
分工:本 skill 管「一条片子从头到尾的顺序」;「某一段画面怎么做」→
cyxj-remotion-overlay;「做完把学到的回写规则」→cyxj-remotion-evolve。三者构成 produce(流程)→ overlay(画面)→ sedimentation(沉淀)闭环,互相指认、不合并。 单一真源纪律:本文件只放阶段顺序 + 判断点 + 指针,不抄任何规则原文(clamp 怎么写、禁什么动画……一律去指针指向的真源读)。真源变了,本文件不用跟着改。
8 阶段编排表
| 阶段 | 做什么 | 判断点 | 真源 |
|---|---|---|---|
| 0 判型 | 定这条片走哪条生产路 | 纯 Remotion 一把成片,还是 Remotion 出图形层 + 后期(剪辑软件)精剪叠层 | 你自己的后期 SOP(本仓不含) |
| 1 起手 | 定起点工程 | 像某个现成项目 → 抄它;否则复制 template/ 开新工程 |
template/CLAUDE.md |
| 2 初始化三查 | 新工程落地后逐项核对 | 依赖钉死不带 ^?跨仓 alias CROSS_REPO_PKGS 配了(template/remotion.config.ts)?zod schema 内联在出片层 src/Root.tsx? |
docs/HARD_RULES.md §7 / §10 / §9 |
| 3 装配 | 拼 props.json 时间轴 |
用注册表选镜头 type(清单以注册表为准,勿写死);每一段画面怎么做(原生造还是复用)→ 整体委托 overlay skill | kit/scenes/sceneMap.ts + cyxj-remotion-overlay |
| 4 字幕 | 按判型走字幕链路 | A 路:Whisper 转写 → 字幕修正 → parseSrt 进片(<SrtCaptions>);B 路:字幕归后期,Remotion 不烧字幕 |
你自己的字幕 SOP(本仓给了 <SrtCaptions> 组件) |
| 5 QA 闸门 | 渲全片前的自检 | ① 新/大改镜头逐个 npm run qa:still -- <type>(见下);② 验收清单逐项勾;③ 用 remotion-rule-reviewer 语义审放行 |
assets/qa-still.mjs |
| 6 渲染交接 | 出成品并交接 | 4K 决策:图形主导 → 1080 合成 + --scale=2;实拍铺满 → 原生 4K 工程。R3F 场景渲染带 --gl=angle |
docs/HARD_RULES.md §10 |
| 7 收工 | 沉淀与进化 | 工程内组件要不要晋升进 kit?本次学到的要不要回写规则?(收工进化闸 hook 会兜底提醒) | kit/CLAUDE.md 组件晋升节 + cyxj-remotion-evolve |
QA 抽帧一键化(阶段 5 的工具)
在出片工程根目录跑(如 template/;工程 package.json 已接 qa:still 入口):
npm run qa:still -- <type> # 默认从 ./props.json 取该 type 的 props/时长,抽入场/中段/落点 3 帧
npm run qa:still -- <type> --frames=a,b,c # 覆盖帧号;更多参数见 assets/qa-still.mjs 头注释
- 脚手架(qa-entry)自动建、自动删,产物落
qa-frames/(自动进.gitignore)。 - 脚本只肯在出片工程根目录跑,在 kit / 仓库根会直接报错退出。
- 为什么必须真渲一帧看:tsc / compositions 全绿也抓不到运行时错;中文尤其要肉眼看(字体没加载好会掉成豆腐块,预览看不出)。
使用姿势
- 用户说「开一条新片」→ 从阶段 0 顺序走;说「渲染 / 出片」→ 从阶段 5 进入,前面阶段只补漏。
- 每进一个阶段,先把该阶段「真源」列里的文件读了再动手;本表只负责「别漏步骤、别乱顺序」。
- 改任何
.tsx/props.json前,官方三查照走(hook 会提醒)。
私有沉淀说明:作者的完整生产线还包含「上传口播一键成片」的前门、后期(剪辑软件)交接 SOP、字幕修正工具链、几十个品牌镜头——这些是私有资产,本仓不含。本 skill 给的是可复用的流程骨架,鼓励你按同样机制沉淀自己的那几层。
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 12d ago First seen · 44 lines · 151 tokens per session scan A 64a2ed2430c1
cyxj-remotion-produce is a skill published in the GitHub repository chenyuxiaojin/cyxj-remotion-starter (6 stars, last pushed 2mo ago), licensed MIT. It adds 151 tokens to every session and 1,384 once invoked, about $0.0008 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
video-strategy-orchestrator
Decides WHEN to use Remotion, HyperFrames, AI video generation, or a hybrid of all three. Use when building video production pipelines, choosing a video tool for a task, or when the user asks 'which video tool should I use?'. Evaluates input type, complexity, determinism needs, and output format to pick the right…
video-director
The Video Production Manager — manages ALL video production under 1ai-content. Delegates to HyperFrames, Remotion, AI video, and FFmpeg based on the task. Has memory via 1ai-hub brain, follows 1ai-rules, and reports to Content Director. Use when the user says 'make a video', 'create video', 'edit video', 'video…
notebook-video
Create complete Chinese 2K warm-ivory engineering-notebook explainer and promotional videos with React, TypeScript and Remotion. The default visual route is pure code-drawn composition (SVG diagrams, mascots, progressive checklists, annotation stickers) synchronized frame-accurately to Chinese TTS word timing, so…
super-video-maker
End-to-end AI video production skill for agentic frameworks. Use when the user asks to make, edit, repurpose, caption, soundtrack, or export videos using HeyGen avatars, Seedance or ByteDance b-roll, OpenAI image generation and editing, AI UGC ads, Remotion, HyperFrames, screen recordings, FFmpeg, captions, ElevenLabs…
vox-director
Turn ONE topic into a finished Vox-style paper-collage explainer / ad video, end to end on the Atlas Cloud API + local ffmpeg — script, collage keyframes, motion, voice-over, music, captions, all automated. Use this whenever the user wants a "Vox style" video, a paper/torn-paper collage animation, a "motion collage"…
stage-edit
Intelligent editing of real user-supplied footage—understand it with transcript/OCR/scene/silence/quality/vision evidence, then choose deterministic timeline operations or a constrained semantic AI edit. Trigger for repurpose, montage, cleanup, localization, narration, or local content changes.