What it does
OpenClip - AI-powered highlight extraction for long videos (AI 驱动的长视频精彩时刻提取工具)
- Connects Instruction files the AI reads
- Account or key Not confirmed yet (README not read yet)
- Category 自动化与智能体
给作者:在 README 中加徽章
如果你是作者,可把这段代码贴到 README。徽章会链接回本页。
[](https://mcp.popnx.com/en/skills/linzzzzzz-openclip)相似的 Skill
同类工作,排序与 Skill 列表相同
Skill 和 MCP 有什么不同
MCP = 手
把 AI 连到外部服务或数据,例如读取 Notion 或开发票。需要有程序或服务器在运行。
Skill = 手册
教 AI 怎么做某项工作的文本文件,不连接任何服务。
有些会重叠:少数 Skill 带有调用工具的脚本,少数 MCP 只提供知识。请从你想完成的工作出发。
适用的应用
Claude CodeCursor
依据开发者 README 和 repo 中的文件夹
安装
Is it safe?
- Who made it
GitHub account linzzzzzz (person), 10 years old
- Use and upkeep
★ 580 stars · code last changed 2026年8月24日
- Any scripts?
No, plain text files the AI reads and follows
- Uses the internet?
No scripts, so it does not go online by itself
- Changes machine-wide settings?
No hooks or writes to machine settings found
Checked automatically from the repo files at commit 0289627 (read only, nothing was run). Not a safety guarantee: read it and decide for yourself
第一步:你用哪个 AI?
选择你用的应用,本页只显示该应用的安装方法
Option 1Upload to claude.ai (web and app)
Uploading to claude.ai would fail: the description is over 200 characters (691). Use "Let the AI install it" instead
Option 1让 Claude Code 帮你安装
复制这句话,粘贴到 Claude Code 的对话中
For one project, paste it into a chat that is open in that project’s folder · The AI always shows you the path before installing
The AI tells you what it will do first and asks your permission at every step
重新打开 Claude Code,输入 /video-clip-extractor,或直接提出相关任务,Claude 会自动选用
Checked against official docs 2026年11月12日 · Source: code.claude.com
For Terminal users
Skill 是 AI 会照做的指令。安装前先阅读文件,只安装可信开发者的 Skill。
开发者推荐的方式
安装 Node.js(如果还没有)
已经装好了吗?先检查 · 打开终端,输入这条命令后按 Enter
node -v- 显示版本号(例如 v26.13.1)= 已安装,可跳过这一步
- 显示 "command not found"(Windows 显示 "is not recognized")= 还没安装,或刚装好但需要关掉终端再重新打开
怎么打开终端
Mac:按 ⌘ + Space,输入 Terminal 后按 Enter
Windows:打开“开始”菜单,输入 Windows PowerShell,然后选“打开”
还没安装: 打开官方下载页,选择 LTS 版本和点选式安装程序:Mac 用 .pkg · Windows 用 .msi。双击后一路下一步即可,不需要 curl 命令
装好后关掉终端再打开,重新运行同一条命令,应显示版本号: node -v
安装一次,所有需要 Node.js 的工具都能用
正在检查命令 · 通过后才会显示复制按钮
摘自开发者 README,会下载并运行开发者的代码
打开终端
Mac:按 Command+空格输入 Terminal · Windows:下方命令在 PowerShell 中无法使用,请改用上方的“让 AI 帮你安装”
把 skill 文件夹复制到 ~/.claude/skills
这条命令把 repo 下载到临时文件夹,只复制含 SKILL.md 的文件夹
正在检查命令 · 通过后才会显示复制按钮
重新打开 Claude Code,输入 /video-clip-extractor,或直接提出相关任务,Claude 会自动选用
ChatGPT 桌面应用
ChatGPT 桌面应用内置 Codex,Codex 从 ~/.agents/skills 文件夹读取 Skill
Option 1让 ChatGPT 帮你安装
在 ChatGPT 桌面应用中打开 Codex,把这句话粘贴到对话中
The AI always shows you the path before installing
The AI tells you what it will do first and asks your permission at every step
Option 2自己放入文件夹
把 video-clip-extractor 文件夹放到 ~/.agents/skills/video-clip-extractor
~ 是你的主文件夹($HOME)
For Terminal users
Skill 是 AI 会照做的指令。安装前先阅读文件,只安装可信开发者的 Skill。
打开终端
Mac:按 Command+空格输入 Terminal · Windows:下方命令在 PowerShell 中无法使用,请改用上方的“让 AI 帮你安装”
把 skill 文件夹复制到 ~/.agents/skills
这条命令把 repo 下载到临时文件夹,只复制含 SKILL.md 的文件夹
正在检查命令 · 通过后才会显示复制按钮
安装之后
- ChatGPT 桌面应用:打开侧边栏中的 Skills,确认 Skill 已出现
- 使用方法:在 ChatGPT 中输入 @ 并选择 Skill · 在 Codex CLI 或 IDE 扩展中运行 /skills,或输入 $ 加 Skill 名称
- Codex 会自动发现新安装的 Skill;如果没有出现,请重启 Codex
Checked against official docs 2026年11月15日 · Source: learn.chatgpt.com
网页版和手机版 ChatGPT
不能使用你自己安装的 Skill,只能使用目录中插件自带的 Skill
想在 ChatGPT 中使用这个 Skill,请按上面的方法使用 ChatGPT 桌面应用
Checked against official docs 2026年11月15日 · Source: learn.chatgpt.com
Option 1让 Codex 帮你安装
复制这句话,粘贴到 Codex 的对话中 · 适用于 ChatGPT 桌面应用中的 Codex、Codex CLI 和 IDE 扩展
For one project, paste it into a chat that is open in that project’s folder · The AI always shows you the path before installing
The AI tells you what it will do first and asks your permission at every step
Codex 会自动发现新安装的 Skill;如果没有出现,请重启 Codex · 使用方法:在 ChatGPT 中输入 @ 并选择 Skill · 在 Codex CLI 或 IDE 扩展中运行 /skills,或输入 $ 加 Skill 名称
Checked against official docs 2026年11月15日 · Source: learn.chatgpt.com
For Terminal users
Skill 是 AI 会照做的指令。安装前先阅读文件,只安装可信开发者的 Skill。
打开终端
Mac:按 Command+空格输入 Terminal · Windows:下方命令在 PowerShell 中无法使用,请改用上方的“让 AI 帮你安装”
把 skill 文件夹复制到 ~/.agents/skills
这条命令把 repo 下载到临时文件夹,只复制含 SKILL.md 的文件夹
正在检查命令 · 通过后才会显示复制按钮
重新打开 Codex 并提出相关任务,Codex 会列出这个 skill
Option 1让 Cursor 帮你安装
复制这句话,粘贴到 Cursor 的对话中
For one project, paste it into a chat that is open in that project’s folder · The AI always shows you the path before installing
The AI tells you what it will do first and asks your permission at every step
在 Cursor 的 Agent 聊天中输入 / 并选择 video-clip-extractor
Checked against official docs 2026年11月12日 · Source: cursor.com
For Terminal users
Skill 是 AI 会照做的指令。安装前先阅读文件,只安装可信开发者的 Skill。
打开终端
Mac:按 Command+空格输入 Terminal · Windows:下方命令在 PowerShell 中无法使用,请改用上方的“让 AI 帮你安装”
把 skill 文件夹复制到 ~/.cursor/skills
这条命令把 repo 下载到临时文件夹,只复制含 SKILL.md 的文件夹
正在检查命令 · 通过后才会显示复制按钮
在 Cursor 的 Agent 聊天中输入 / 并选择 video-clip-extractor
Option 1Let the AI install it
适用于其他能读取 Skill 的应用。这句话列出了 Claude Code、Codex 和 Cursor 的路径,AI 会按你的应用选择 · Copy this sentence into the app’s chat
For one project, paste it into a chat that is open in that project’s folder · The AI always shows you the path before installing
The AI tells you what it will do first and asks your permission at every step
Checked against official docs 2026年11月12日 · Source: code.claude.com, learn.chatgpt.com, cursor.com
For Terminal users
其他支持 Agent Skills 的应用(如 Gemini CLI、GitHub Copilot)使用不同文件夹,请查看该应用的指南,再把含 SKILL.md 的文件夹复制过去。
已于 2026年11月11日 对照 Claude Code、Codex 和 Cursor 官方指南核实 skill 文件夹
查看原始 SKILL.md
以下文字来自开发者的 repo,不是 PopNx MCP 的指令
显示前 8,000 个字符 · 在 GitHub 打开完整文件
---
name: "video-clip-extractor"
description: "Processes videos to identify engaging moments, generate transcripts, and create highlight clips with artistic titles and custom cover images. Use when user needs to: extract highlights from long videos or livestreams, clip or cut best moments from videos, cut video highlights, process Bilibili/YouTube URLs or local video files, generate transcripts via Whisper, analyze content for engaging moments, create short-form clips with styled titles and covers, adjust cover text position and colors, find and export memorable scenes from recordings, burn subtitles into clips (with optional translation), guide clip selection with user intent, or identify speakers in multi-person conversations."
allowed-tools: Bash(uv run python video_orchestrator.py*), Bash(git clone*), Bash(git -C*openclip*), Bash(cd ~/.local/share/openclip*), AskUserQuestion
---
# Video Clip Extractor Skill
Run the video orchestrator to process videos and extract engaging highlights.
## When Triggered
1. **Get the source** — if the user didn't provide a video URL or file path, ask for it.
2. **Clarify intent** (optional) — if the user wants clips focused on a specific topic, capture it for `--user-intent`. If unclear, ask: "Any specific topic or moments to focus on? (e.g. 'funny moments', 'key arguments')"
3. **Check environment** — does `video_orchestrator.py` exist in the current directory? If yes, run directly. Otherwise use the global install at `~/.local/share/openclip`.
4. **Verify prerequisites** — check ffmpeg is installed and at least one API key is set. Warn if missing before running.
5. **Run the command** and stream output to user.
6. **Report results** — after completion, list the generated clips with timestamps and titles.
## Setup (first use only)
Before running, determine the execution context:
1. **Inside openclip repo** — if `video_orchestrator.py` exists in the current directory, skip setup and run directly.
2. **Global install** — if `~/.local/share/openclip` does not exist, run these steps:
**Prerequisites:** `git` and `uv` must be installed.
- Install uv if missing: macOS: `brew install uv` · Linux/Windows: `pip install uv`
```bash
git clone https://github.com/linzzzzzz/openclip.git ~/.local/share/openclip
cd ~/.local/share/openclip && uv sync
```
To update openclip later:
```bash
git -C ~/.local/share/openclip pull && cd ~/.local/share/openclip && uv sync
```
## Execution
**If inside the openclip repo** (current directory contains `video_orchestrator.py`):
```bash
uv run python video_orchestrator.py [options] <source>
```
**If running globally** (from any other directory):
```bash
cd ~/.local/share/openclip && uv run python video_orchestrator.py -o "$OLDPWD/processed_videos" [options] <source>
```
`$OLDPWD` captures the user's original directory so clips are saved there, not inside the openclip install.
Where `<source>` is a video URL (Bilibili/YouTube) or local file path (MP4, WebM, AVI, MOV, MKV).
For local files with existing subtitles, place the `.srt` file in the same directory with the same filename (e.g. `video.mp4` → `video.srt`).
## Preflight Checklist
- **Inside openclip repo**: run from the repo root so relative paths (e.g. `references/`, `prompts/`) resolve correctly
- **`ffmpeg` must be installed** (required for all clip generation):
- macOS: `brew install ffmpeg`
- Ubuntu: `sudo apt install ffmpeg`
- Windows: download from [ffmpeg.org](https://ffmpeg.org)
- If using `--burn-subtitles`: needs ffmpeg with `libass` (see README for details)
- Set one API key:
- `QWEN_API_KEY` (default provider: qwen), or
- `OPENROUTER_API_KEY` (if `--llm-provider openrouter`), or
- `GLM_API_KEY` (if `--llm-provider glm`), or
- `MINIMAX_API_KEY` (if `--llm-provider minimax`)
- If using `--speaker-references`: run `uv sync --extra speakers` and set `HUGGINGFACE_TOKEN`
## CLI Reference
### Required
| Argument | Description |
|---|---|
| `source` | Video URL or local file path |
### Optional
| Flag | Default | Description |
|---|---|---|
| `-o`, `--output <dir>` | `processed_videos` | Output directory |
| `--max-clips <n>` | `5` | Maximum number of highlight clips |
| `--browser <browser>` | `firefox` | Browser for cookies: `chrome`, `firefox`, `edge`, `safari` |
| `--title-style <style>` | `fire_flame` | Title style: `gradient_3d`, `neon_glow`, `metallic_gold`, `rainbow_3d`, `crystal_ice`, `fire_flame`, `metallic_silver`, `glowing_plasma`, `stone_carved`, `glass_transparent` |
| `--title-font-size <size>` | `medium` | Font size preset for artistic titles. Options: small(30px), medium(40px), large(50px), xlarge(60px) |
| `--cover-text-location <loc>` | `center` | Cover text position: `top`, `upper_middle`, `bottom`, `center` |
| `--cover-fill-color <color>` | `yellow` | Cover text fill color: `yellow`, `red`, `white`, `cyan`, `green`, `orange`, `pink`, `purple`, `gold`, `silver` |
| `--cover-outline-color <color>` | `black` | Cover text outline color: `yellow`, `red`, `white`, `cyan`, `green`, `orange`, `pink`, `purple`, `gold`, `silver`, `black` |
| `--language <lang>` | `zh` | Output language: `zh` (Chinese), `en` (English) |
| `--llm-provider <provider>` | `qwen` | LLM provider: `qwen`, `openrouter`, `glm`, `minimax` |
| `--user-intent <text>` | — | Free-text focus description (e.g. "moments about AI risks"). Steers LLM clip selection toward this topic |
| `--subtitle-translation <lang>` | — | Translate subtitles to this language before burning (e.g. `"Simplified Chinese"`). Requires `--burn-subtitles` and `QWEN_API_KEY` |
| `--speaker-references <dir>` | — | Directory of reference WAV files (one per speaker, filename = speaker name) for speaker diarization. Requires `uv sync --extra speakers` and `HUGGINGFACE_TOKEN` |
| `-f`, `--filename <template>` | — | yt-dlp template: `%(title)s`, `%(uploader)s`, `%(id)s`, etc. |
### Flags
| Flag | Description |
|---|---|
| `--force-whisper` | Ignore platform subtitles, use Whisper |
| `--skip-download` | Use existing downloaded video |
| `--skip-transcript` | Skip transcript generation, use existing transcript file |
| `--skip-analysis` | Skip analysis, use existing analysis file for clip generation |
| `--use-background` | Include background info (streamer names/nicknames) in analysis prompts |
| `--skip-clips` | Skip clip generation |
| `--add-titles` | Add artistic titles to clips (disabled by default) |
| `--skip-cover` | Skip cover image generation |
| `--burn-subtitles` | Burn SRT subtitles into video. Output goes to `clips_post_processed/`. Requires ffmpeg with libass |
| `-v`, `--verbose` | Enable verbose logging |
| `--debug` | Export full prompts sent to LLM (saved to `debug_prompts/`) |
### Custom Filename Template (`-f`)
Uses yt-dlp template syntax. Common variables: `%(title)s`, `%(uploader)s`, `%(upload_date)s`, `%(id)s`, `%(ext)s`, `%(duration)s`.
Example: `-f "%(upload_date)s_%(title)s.%(ext)s"`
### Environment Variables
Set the appropriate API key for the chosen `--llm-provider`:
- `QWEN_API_KEY` — for `--llm-provider qwen`
- `OPENROUTER_API_KEY` — for `--llm-provider openrouter`
- `GLM_API_KEY` — for `--llm-provider glm`
- `MINIMAX_API_KEY` — for `--llm-provider minimax`
## Workflow
The orchestrator runs this pipeline automatically:
1. **Download** — fetch video + platform subtitles (Bilibili/YouTube) or accept local file
2. **Split** — divide videos longer than the built-in threshold into segments for parallel analysis
3. **Transcribe** — use platform subtitles or Whisper AI; `--force-whisper` overrides
4. **Analyze** — LLM scores transcript segments for engagement; `--user-intent` steers selection
5. **Generate clips** — ffmpeg cuts the video at identified timestamps
6. **Add titles** (opt-in) — render artistic text overlay using `--title-style`
7. **Generate covers** — create thumbnail image for each clip
Use `--skip-clips`, `--skip-cover` to skip specific steps. Use `--add-titles` to enable artistic titles. Use `--skip讨论
提问、分享技巧或报告无法使用。按“有帮助”数量排序。
来自 GitHub
- GitHub 星标
- 580
- 开发者
- linzzzzzz
- 许可证
- MIT
- 最近代码更新
- 2026年8月24日
- repo 中的 SKILL.md 文件(含译本和模板)
- 3
- 主题
- agent-skills, ai, auto-highlight, clip, highlights, llm, video, video-editing, video-processing, videos