← Back to Skills Marketplace
84
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install clipping-video-highlights
Description
长视频（YouTube链接/本地视频）→ N个短高光片段+烧字幕+AI封面。100%免费方案：yt-dlp字幕提取+MiniMax LLM选高光+MiniMax图片生成封面+ffmpeg剪辑。
README (SKILL.md)
长视频高光剪辑技能（免费方案）
零成本复刻Clawvard clipping-video-highlights，不花一分钱。
适用场景
- YouTube视频 → 抖音/小红书高光片段
- 播客/讲座/直播回放 → 短高光剪辑
- 本地视频批量剪辑
核心流程（4步）
Step 1: yt-dlp --write-subs → 自动提取YouTube字幕（免费）
Step 2: MiniMax LLM → 从字幕文本选高光时间点（已有配额）
Step 3: MiniMax图片生成 → AI生成封面（已有配额）
Step 4: ffmpeg剪辑+烧字幕 → 输出成品短视频
使用方法
python3 clipper.py \x3CYouTube_URL> \x3C输出目录> [片段数量]
# 示例：
python3 clipper.py "https://www.youtube.com/watch?v=xxx" "./output" 5
依赖
ffmpeg/ffprobe— 视频处理yt-dlp— 字幕提取- MiniMax API（已有：
sk-cp-qLf6tET6ParE9D35M3O0_5TE） - 腾讯COS（已有配置）
输出
cover.jpg— AI生成封面clip_01.mp4,clip_02.mp4, ... — 剪辑成品（含烧字幕）
Usage Guidance
Review the MiniMax credential situation before installing or running this skill. Prefer a version that removes embedded keys, declares required tools and credentials, and clearly explains that transcript text is sent to MiniMax. Use a dedicated output folder and avoid private videos unless you accept the external provider data flow.
Capability Analysis
Type: OpenClaw Skill
Name: clipping-video-highlights
Version: 1.0.0
The skill bundle contains a hardcoded MiniMax API key (sk-cp-qLf6tET6ParE9D35M3O0_5TE) and claims to be a 'free replica' of a paid service, which is a common indicator of grayware or potentially compromised tools. The script clipper.py relies on an external, non-standard CLI utility named 'mmx' for LLM and image generation tasks, which is not a standard dependency and its origin is unverified. Additionally, the documentation in SKILL.md mentions Tencent COS configurations that are entirely absent from the provided code, suggesting either incomplete implementation or undisclosed external interactions.
Capability Assessment
Purpose & Capability
The main YouTube-to-clips workflow is coherent with the description, but the artifacts also claim local video and Tencent COS support while the code only implements YouTube subtitle/section downloading and does not use COS.
Instruction Scope
SKILL.md presents a MiniMax API key/quota as already available, while registry metadata declares no credentials; users cannot tell whose account is used or what scope applies.
Install Mechanism
There is no install spec and metadata declares no required binaries, but the code requires ffmpeg, ffprobe, yt-dlp, and an mmx CLI; package.json also contains an unused npm dependency.
Credentials
External provider use is purpose-aligned for AI highlight selection and cover generation, but transcript sharing and MiniMax credential handling are not clearly declared or bounded.
Persistence & Privilege
No background persistence or self-starting behavior is shown; privilege risk is mainly the undeclared provider/API credential use rather than local persistence.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install clipping-video-highlights - After installation, invoke the skill by name or use
/clipping-video-highlights - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Initial release
Metadata
Frequently Asked Questions
What is clipping-video-highlights?
长视频（YouTube链接/本地视频）→ N个短高光片段+烧字幕+AI封面。100%免费方案：yt-dlp字幕提取+MiniMax LLM选高光+MiniMax图片生成封面+ffmpeg剪辑。 It is an AI Agent Skill for Claude Code / OpenClaw, with 84 downloads so far.
How do I install clipping-video-highlights?
Run "/install clipping-video-highlights" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is clipping-video-highlights free?
Yes, clipping-video-highlights is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does clipping-video-highlights support?
clipping-video-highlights is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created clipping-video-highlights?
It is built and maintained by freak30 (@freak30); the current version is v1.0.0.
More Skills