Xunyu hand drawn healing story
Agent Skills collection — hand-drawn shorts, flight-monitor-agent, hotel-watch-agent and more
npx -y skills add xunyuAI/skills --skill xunyu-hand-drawn-healing-storyAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
- 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
手绘视频生成（xunyu）：自然语言一句话即可产出「白色极简手绘（通用）」竖屏短视频，主题不限。 顶部叠字、左→右 wipe、可选黑白→淡彩，默认静音，默认引擎 video-use（ffmpeg）； 另支持 Remotion / Hyperframes / HTML。支持 mode=smoke（占位图快速测通）与 mode=full。 含 10 个通用示例主题池（可自拟覆盖）、完整流水线、四引擎、生图通道与 BGM 可选能力。 Use whenever the user mentions 手绘视频、手绘竖屏、白色极简手绘、横扫上色、通用手绘短视频、 视频号/抖音无配音叠字短视频、hand-drawn vertical shorts、xunyu 手绘、 或要从示例主题池选题 / 自拟题材出片—even if they do not name this skill. Prefer this over generic video-use or handdraw-story-video for this white-minimal hand-drawn vertical format.
SKILL.md
8.5 KB, ~2.6k tokens by cl100k_base, as published. Nobody here has run it
xunyu · 手绘短视频（通用）
稳定 ID：xunyu-hand-drawn-healing-story。对外可称「手绘视频生成」。
视觉规格：白色极简手绘（通用）——主题不限，用户说什么题材就做什么。成片默认静音（BGM none）。
文档分层
- 本文件：自然语言入口 + Smoke 决策树（短）。
- 完整能力与逐步出片 →
[references/pipeline.md](references/pipeline.md)- 四引擎命令与分支 →
[references/engines.md](references/engines.md)- 选型对比（优缺点）→
[README.md](README.md)
最小可测闭环（Smoke path）— 先读这个
目标：用户一句话 → 能出一条短 mp4（或明确交付物路径）。
| 步 | 做什么 | 默认 |
|---|---|---|
| 1 | 解析用户话 → 定 theme_id | 未提主题 → 02 时间与等待 |
| 2 | 写分镜 md + shots.json | 6–9 镜（shot_count，默认 7） |
| 3 | 生图 | mode: smoke → 跳过生图，用占位图 / 已有素材目录；full 才真生图 |
| 4 | 合成 | 默认 video-use；无素材时 scripts/smoke_render.sh 也能出短 mp4 |
| 5 | 告诉用户 | 一行输出路径（exports/*.mp4 或预览目录） |
# 集目录内一键 smoke（最少依赖：ffmpeg + Python/PIL）
# 安装后在 skill 根目录执行：
bash scripts/smoke_render.sh
# 或全局路径（视 Agent 安装位置而定）：
# bash ~/.agents/skills/xunyu-hand-drawn-healing-story/scripts/smoke_render.sh
自然语言入口（最多问 1 个问题）
用户说类似下面任意一句，直接跑，不要盘问：
- 「做一条手绘短视频」
- 「做一条 50 秒竖屏手绘视频，主题选时间与等待」
- 「用职场通勤题材做一条白色极简手绘短视频」
- 「大概一分钟的手绘短视频」
缺省规则（不问也行）
| 缺什么 | 用什么 |
|---|---|
| 主题 | theme_id: "02"（时间与等待） |
| 时长 | ~50s（target_duration_sec: 50）；说「一分钟」→ 55–60s |
| 画幅 | 9:16 · 1080×1920 · 30fps |
| 引擎 | video-use |
| BGM | none |
| 模式 | 无 key / 用户说「测试」「先跑通」→ mode: smoke；否则 full |
| 镜数 | 6–9（默认 7） |
唯一可问的 1 个问题（仅当关键信息互相矛盾时）
例：同时说「只要预览」又「必须出 mp4」。否则 0 问，直接 defaults 闭环。
禁止开跑时追问生图通道、wipe 毫秒、四引擎对比——这些见 references/。
能力一览（细节下沉 references）
| 能力 | 默认 / 说明 | 详文 |
|---|---|---|
| 自然语言出片 | 最多 1 问 + defaults | 本文件 |
| Smoke / Full | smoke 占位测通；full 真生图 | [pipeline.md](references/pipeline.md) |
| 10 示例主题（可自拟） | 01–10；默认 02；任意题材可覆盖 | [themes.md](references/themes.md) |
| 四引擎 | 默认 video-use；另 html / remotion / hyperframes | [engines.md](references/engines.md) · [README.md](README.md) |
| 生图通道 | auto / 内置 / OpenAI / http_api / skip | [image-gen.md](references/image-gen.md) |
| 叠字字体 | OFL 站酷快乐体；无参考 PNG | [typography.md](references/typography.md) |
| BGM | 默认 none；可选 local / platform | [bgm.md](references/bgm.md) |
| 全量配置 | assets/defaults.yaml + 释义 | [defaults.md](references/defaults.md) |
| 高级调参 / 只出分镜 | wipe、镜数、engine 覆盖 | [advanced.md](references/advanced.md) |
| 发布自检 | 短清单 | [release-checklist.md](references/release-checklist.md) |
| 多 Agent | dbs-bridge 软链真源 | 文末 bridge |
对话脚本（用户说了什么 → 你做什么）
A. 「做一条手绘短视频」/ 「大概一分钟」
mode: smoke（除非环境已有生图且用户要成品画质）- 选题默认
02；写 7 镜分镜 +shots.json+ 双平台标题 bash scripts/smoke_render.sh（或项目内等价调用）- 回复一行：
成片：…/exports/xxx.mp4
B. 「主题用自我和解，engine 用 html 只要预览」
theme_id: "06"，engine: html- 写分镜 + 填
templates/html→open index.html - 不跑 ffmpeg
C. 「wipe 慢一点，出分镜不要渲染」
- 只写分镜 +
shots.json+config.yaml（wipe.duration_sec: 1.4） - 停。不生图、不合成
D. 明确要成片 / 有图 / 有 API key
mode: full→ 按[pipeline.md](references/pipeline.md)+[image-gen.md](references/image-gen.md)生图（成对 bw+color）make_captions.py→render_ffmpeg.sh（或其它引擎）- 一行输出路径 + 双平台标题
默认摘要
| 项 | 默认 |
| ------ | ----------------------------------------------------- | ---------- | --------------- |
| 视觉规格 | 白色极简手绘（通用 · 主题不限） |
| 示例主题 | references/themes.md（01–10，可自拟覆盖） |
| 引擎 | video-use（html | remotion | hyperframes） |
| 字体 | OFL 站酷快乐体 assets/fonts/ZCOOLKuaiLe-Regular.ttf |
| 参考图 | 无（skill 不附带对标 PNG） |
| BGM | none |
| 生图 | image_gen.channel: auto；smoke 可跳过 |
机器默认：assets/defaults.yaml。项目 config.yaml 浅合并覆盖。密钥不进仓库。
按需阅读
| 需要 | 文件 |
|---|---|
| 完整流水线（选题→质检） | [references/pipeline.md](references/pipeline.md) |
| 全部配置项释义 | [references/defaults.md](references/defaults.md) |
| 选题 | [references/themes.md](references/themes.md) |
| 生图通道 | [references/image-gen.md](references/image-gen.md) |
| 四引擎实现 | [references/engines.md](references/engines.md) |
| 四引擎优缺点对比 | [README.md](README.md) |
| 叠字 | [references/typography.md](references/typography.md) |
| BGM（none / local / platform） | [references/bgm.md](references/bgm.md) |
| 高级调参 | [references/advanced.md](references/advanced.md) |
| 发布自检 | [references/release-checklist.md](references/release-checklist.md) |
引擎一览（细节下沉）
engine | 动作 |
|---|---|
| video-use（默认） | make_captions.py → render_ffmpeg.sh；smoke 用 smoke_render.sh |
html | 填 templates/html，浏览器预览，不强制 mp4 |
remotion / hyperframes | 见 [references/engines.md](references/engines.md) + templates/ |
输出目录
白色极简手绘/{集号}-{短标题}/
├── config.yaml
├── {集号}-分镜.md
├── shots.json
├── {集号}-素材/ # smoke 可为占位图
├── out/
└── exports/{slug}.mp4
交付时必须用一行写出最终路径。
边界
- 本 skill：白底极简手绘（主题通用）· 叠字 · wipe+上色 · 四引擎（默认 video-use）· NL 优先 + smoke。
handdraw-story-video：好事八拍模具——勿混。hand-drawn-styles：只出画风 prompt。- 通用
video-use：底层能力；白色极简手绘竖屏规格以本 skill 为准。
安装（skills.sh / GitHub）：
npx skills add sxyfe/skills@xunyu-hand-drawn-healing-story -g -y
What ships with it: 45 files
1800.4 KB alongside SKILL.md, 6 of them executable
assets/
- defaults.yaml4.7 KB
- fonts/.gitkeep0 B
- fonts/OFL-ZCOOLKuaiLe.txt4.3 KB
- fonts/README.md517 B
- fonts/RoundedHandwrite.ttf23 B
- fonts/ZCOOLKuaiLe-Regular.ttf1479.5 KB
- readme/engines.svg3.2 KB
- readme/hero.svg3.6 KB
- readme/section-difference.svg1.1 KB
- readme/showcase.png215.3 KB
- readme/workflow.svg2.2 KB
evals/
- evals.json3.6 KB
references/
- advanced.md3.2 KB
- bgm.md3.6 KB
- defaults.md7.4 KB
- engines.md5.5 KB
- image-gen.md3.1 KB
- pipeline.md9.6 KB
- release-checklist.md481 B
- themes.md5.3 KB
- typography.md1.7 KB
scripts/
- make_captions.pyruns3.0 KB
- render_ffmpeg.shruns3.2 KB
- smoke_render.shruns3.0 KB
templates/
- html/assets/.gitkeep4.0 KB
- html/assets/README.md85 B
- html/index.html1.1 KB
- html/main.jsruns3.8 KB
- html/style.css2.5 KB
- hyperframes/assets/README.md101 B
- hyperframes/hyperframes.json265 B
- hyperframes/index.html3.4 KB
- hyperframes/meta.json115 B
- hyperframes/package.json525 B
- README.md648 B
- remotion/package.json527 B
- remotion/public/README.md104 B
- .gitignore105 B
- LICENSE1.1 KB
- README.md10.8 KB
5 more files not listed here. See all 45 in the repository.
Gives 0 of the 12 instructions most video audio skills give in ~2.6k tokens
Counted across 622 of the 795 authors here whose files we hold, read 2026-08-07
- Read individual rule files for detailed explanationsin 21 of 622, across 10 files
- Render final videoin 13 of 622, across 6 files
- Use WAV PCM 16kHz mono audio formatin 12 of 622, across 3 files
- Use this skill when dealing with Remotion codein 11 of 622, across 4 files
- Save generated audio to a WAV filein 11 of 622, across 4 files
- Handle conversion errors gracefullyin 10 of 622, across 6 files
- Add captions to videos alwaysin 10 of 622, across 4 files
- Generate music from text descriptions using MusicGenin 9 of 622, across 2 files
- Do not skip pipeline layersin 9 of 622, across 3 files
- Do not make one tool do everythingin 9 of 622, across 3 files
- Use Azure Document Intelligence for complex PDFsin 9 of 622, across 4 files
- Never ask the user to paste their full API keyin 9 of 622, across 3 files
Said here and by no other author read
- Run defaults loop without asking questions
- Ask at most one question when information conflicts
- Use default theme 02 if unspecified
- Generate a 6 to 9 shot storyboard
- Use smoke mode for initial tests
- Execute smoke render script to produce video
Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.