菜单
Create a short digest/story video from raw footage using Henshusha's manual edit plan, Timeline JSON, FFmpeg render, and optional Remotion preview/render path.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/eiei114/henshusha --skill henshusha-digest-story命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
基于 SOC 职业分类
正在显示 SKILL.md
| name | henshusha-digest-story |
| description | Create a short digest/story video from raw footage using Henshusha's manual edit plan, Timeline JSON, FFmpeg render, and optional Remotion preview/render path. |
Use when the user wants to quickly make a digest, highlight reel, story cut, recap, short video, or editing demo from existing footage.
Make editing feel simple and visually satisfying on the first run:
raw video → 3-5 story beats → scripts/digest-story.md → scripts/edit-plan.md → timelines/main.timeline.json with Remotion text presets → dry-run → FFmpeg MP4 and Remotion preview/render.
projects/<project-name>/ (default: projects/sample-video).npm run doctor:updates when working inside a generated workspace and network access is acceptable. If an update is available, tell the user before editing; do not block the digest if the check fails.sources/raw/.scripts/digest-story.md with 3-5 beats:
scripts/edit-plan.md:
headline-popkaraoke-caption when timing/progress helpsbottom-captionlower-thirdquote-cardaccent, label, or speaker when it clarifies the beattimelines/main.timeline.json using only manual timings.henshusha validate projects/<project-name>.henshusha render projects/<project-name> --dry-run and inspect jobs/render-plan.json.henshusha remotion-props projects/<project-name> so Remotion can preview the same story timeline.henshusha render projects/<project-name> to produce renders/output.mp4.npm run remotion:previewnpm run remotion:renderUse this shape in scripts/digest-story.md:
# Digest Story
## Intent
- Audience:
- Promise:
- Target length:
## Beats
1. Hook — 0:00-0:02 — `headline-pop` title — why watch now
2. Context — 0:02-0:04 — `karaoke-caption` caption — setup
3. Key moment — 0:04-0:07 — `bottom-caption` caption — strongest evidence
4. Payoff — 0:07-0:09 — `lower-third` or `quote-card` — lesson or reveal
5. Outro — 0:09-0:10 — `bottom-caption` caption — closing
## Source ranges
- raw/source.mp4 00:00:00-00:00:02 -> Hook
- raw/source.mp4 00:00:02-00:00:04 -> Context
- raw/source.mp4 00:00:04-00:00:07 -> Key moment
- raw/source.mp4 00:00:07-00:00:09 -> Payoff
- raw/source.mp4 00:00:09-00:00:10 -> Outro
Use these fields on title / caption track items whenever possible:
{ "start": 0, "end": 2, "text": "素材から\nストーリーへ", "preset": "headline-pop", "label": "HOOK", "accent": "#ffcf33" }
Available starter presets:
headline-pop / bold-center: opening hook or big transition.bottom-caption: readable default subtitle.karaoke-caption: short caption with a progress underline.lower-third: named payoff card; use speaker when useful.quote-card: centered quote / memorable conclusion.If unsure, use: Hook = headline-pop, middle captions = bottom-caption or karaoke-caption, payoff = lower-third, outro = quote-card or bottom-caption.