跳至正文
OpenAgentSkill
浏览已开放 · 目录记录日期 2026-11-21 · 自动更新和账户服务暂时关闭。
其他技能

clip-cutter

Cut a picked moment into a finished 1080x1920 vertical clip with word-by-word captions, and a hook box. Use after moment-finder, or when the user says "cut it", "make the clip", "make it vertical".

Bodila51 · 已有静态检查记录

安装与使用

运行命令前先阅读来源与依赖说明。复制命令不会自动安装或执行。

npx skills add Bodila51/gtavi-claudecode --skill clip-cutter

技能说明


name: clip-cutter description: Cut a picked moment into a finished 1080x1920 vertical clip with word-by-word captions, and a hook box. Use after moment-finder, or when the user says "cut it", "make the clip", "make it vertical".

Clip cutter

Run:

python "${CLAUDE_PLUGIN_ROOT}/scripts/cut.py" jobs/<id> --pick <n> --layout fill|blur|facecam --hook "TEXT"

  • fill: crop the game video to fill the frame. Default for gameplay.
  • blur: full video centered over a blurred copy. Use when cropping would cut off the action.
  • facecam: gameplay with the streamer's camera shown. Use for reactions.

Before running:

  1. Get 3 hook options from the hook-writer skill and use the best one.

After running:

  1. Output is an mp4 in media/edits. Check it is 1080x1920, the hook shows in the first 2 seconds, captions are readable. No credit text is burned into the video; the poster credits the creator in the post.
  2. Run rule-check on it. Fix and re-cut until it passes.
  3. Trim 7 to 30 seconds for most clips. End on the payoff.
  4. Show the human the file, and the caption text, and ask for "yes". Do not upload anything.