← 返回 Skills 市场
mayank8290

Clips Machine

作者 Mayank8290 · GitHub ↗ · v1.2.1
cross-platform ✓ 安全检测通过
2131
总下载
7
收藏
1
当前安装
4
版本数
在 OpenClaw 中安装
/install clips-machine
功能描述
Transform long videos into viral short-form clips. Auto-detect best moments, add trendy captions, export for TikTok/Reels/Shorts. Self-contained, no external modules. 100% free tools.
使用说明 (SKILL.md)

Clips Machine

Transform long videos into viral short-form clips. Auto-detect the best moments, add trendy captions, export for TikTok/Reels/Shorts.

100% FREE tools - Runs entirely on your machine.

Love this skill? Support the creator and help keep it free: Buy Me a Coffee

What This Skill Does

  1. Input any long video (YouTube URL, podcast, stream, local file)
  2. Transcribe with timestamps using Whisper (free, local)
  3. Detect viral-worthy moments using AI analysis
  4. Cut the best 30-60 second segments
  5. Caption with animated, trendy text styles
  6. Export in vertical 9:16 format ready for upload

Quick Start

Turn this podcast into viral clips: https://youtube.com/watch?v=xyz

Extract the 5 best moments from my-interview.mp4 and add captions

Commands

From YouTube URL

/clips-machine https://youtube.com/watch?v=VIDEO_ID

From Local File

/clips-machine /path/to/video.mp4

Custom Number of Clips

/clips-machine VIDEO --clips 10

Caption Styles

/clips-machine VIDEO --style [style]

Available styles:

  • hormozi - Alex Hormozi style (bold, word-by-word highlight) - Most Viral
  • minimal - Clean white text
  • karaoke - Word-by-word color change
  • news - Lower third style
  • meme - Impact font, top/bottom

How Viral Detection Works

The AI analyzes the transcript looking for:

  1. Hook Potential - Strong opening statements
  2. Emotional Peaks - Passion, excitement, surprise
  3. Quotable Lines - Memorable one-liners
  4. Controversial Takes - Debate-worthy opinions
  5. Surprising Facts - "Did you know" moments
  6. Actionable Advice - Clear takeaways

Each moment gets a "virality score" from 1-100.

Output Structure

~/Videos/OpenClaw/clips-[video-name]/
├── transcript.json      # Full transcript with timestamps
├── viral_moments.json   # Detected moments with scores
├── clip_001.mp4         # First viral clip (vertical, captioned)
├── clip_002.mp4         # Second viral clip
├── clip_003.mp4         # ...
└── summary.md           # Overview of all clips

Supported Sources

Source Example
YouTube https://youtube.com/watch?v=...
TikTok https://tiktok.com/@user/video/...
Twitter/X https://twitter.com/user/status/...
Twitch VOD https://twitch.tv/videos/...
Local MP4 /path/to/file.mp4

Requirements

  • FFmpeg (brew install ffmpeg)
  • yt-dlp (brew install yt-dlp)
  • Whisper.cpp (brew install whisper-cpp)

Setup

# Install dependencies
brew install ffmpeg yt-dlp whisper-cpp

# Or on Linux
sudo apt install ffmpeg
pip install yt-dlp
# Build whisper.cpp from source

Monetization

Method Potential
Clip service for creators $50-150/video
Monthly retainer $500-2,000/client
Podcast clipping agency $2,000-5,000/mo
Sell this skill $100-300 on ClawHub

Examples

Podcast to Clips

Take this 2-hour podcast and find the 10 best moments:
https://youtube.com/watch?v=PODCAST_ID
Make them Hormozi-style with bold captions.

Interview Highlights

/clips-machine interview.mp4 --clips 5 --style minimal

Gaming Stream

Extract funny moments from my Twitch VOD:
https://twitch.tv/videos/12345
Add meme-style captions

Support This Project

If this skill saved you time or made you money, consider buying me a coffee!

Buy Me a Coffee

Every coffee helps me build more free tools for the community.

Built for OpenClaw | Powered by Whisper + FFmpeg | Support the Creator

安全使用建议
This skill appears coherent with its stated purpose, but before installing/run it: (1) ensure ffmpeg, yt-dlp and whisper-cpp are obtained from trusted sources and installed in a secure environment; (2) confirm you have rights to download/process any remote video you give it; (3) review or run the script in a sandbox/VM if you're concerned about executing third-party code; (4) note it writes outputs to ~/Videos/OpenClaw (or OUTPUT_DIR if set) and will download remote media—monitor disk/network usage. If you want higher assurance, review the full clips_machine.py file (the visible portions are standard) or run it on a non-sensitive machine first.
功能分析
Type: OpenClaw Skill Name: clips-machine Version: 1.2.1 The OpenClaw AgentSkill 'clips-machine' is classified as benign. The `clips_machine.py` script demonstrates good security practices by using `subprocess.run` with a list of arguments (avoiding `shell=True`), implementing URL validation against a whitelist of allowed domains to prevent SSRF, and using `tempfile.mkstemp` for secure temporary file creation. Additionally, the script sanitizes text before generating ASS subtitle files to mitigate injection risks. The `SKILL.md` documentation is clear, aligns with the script's functionality, and contains no evidence of prompt injection attempts against the OpenClaw agent or instructions for malicious behavior. All external dependencies (`ffmpeg`, `yt-dlp`, `whisper-cpp`) are legitimate tools for video processing.
能力评估
✓ Purpose & Capability
Name/description (video -> short clips) match the actual behavior: it invokes yt-dlp, ffmpeg, and whisper-cpp to download, transcribe, cut, crop, caption and export clips. The declared required binaries are appropriate and necessary for the stated functionality.
✓ Instruction Scope
SKILL.md and the script restrict actions to downloading permitted domains, local transcription, clip detection, cutting, captioning, and exporting into a local output folder. No instructions ask the agent to read unrelated system files, secrets, or to send data to unknown endpoints.
✓ Install Mechanism
There is no install spec (instruction-only install), and the script itself uses only standard subprocess calls to system binaries. The user must install ffmpeg, yt-dlp and whisper-cpp manually as documented; no risky third-party download/execute steps are embedded.
✓ Credentials
The skill declares no required environment variables or credentials. The code optionally reads OUTPUT_DIR (to set an output path) and resolves local whisper model paths; no secrets or unrelated credentials are requested.
✓ Persistence & Privilege
always is false and the skill does not request persistent system-wide changes or modify other skills' configurations. It runs as a standalone script and writes outputs only to a user-writable directory.
如何使用
  1. 确保已安装 OpenClaw（本地或 Docker 部署）
  2. 在对话框中输入安装命令：/install clips-machine
  3. 安装完成后，直接呼叫该 Skill 的名称或使用 /clips-machine 触发
  4. 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.2.1
Added support link at top
v1.2.0
Added support/donation link
v1.1.0
v1.1.0: Self-contained (no external modules), security hardened (SSRF protection, secure tempfiles, URL validation, ASS injection prevention), fixed all scanner flags
v1.0.0
Initial release: Effortlessly transform long videos into viral short-form clips. - Auto-detects the best moments using AI with virality scoring. - Adds animated, trendy captions in multiple styles (Hormozi, minimal, meme, more). - Supports direct export for TikTok, Reels, and Shorts in vertical format. - Accepts YouTube, TikTok, Twitter/X, Twitch VOD, and local MP4 inputs. - 100% free and runs fully on your local machine (requires ffmpeg, yt-dlp, whisper-cpp).
元数据
Slug clips-machine
版本 1.2.1
许可证 —
累计安装 1
当前安装数 1
历史版本数 4
常见问题

Clips Machine 是什么？

Transform long videos into viral short-form clips. Auto-detect best moments, add trendy captions, export for TikTok/Reels/Shorts. Self-contained, no external modules. 100% free tools. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 2131 次。

如何安装 Clips Machine？

在 OpenClaw 或 Claude Code 对话框中运行命令「/install clips-machine」即可一键安装，无需额外配置。

Clips Machine 是免费的吗？

是的，Clips Machine 完全免费（开源免费），可自由下载、安装和使用。

Clips Machine 支持哪些平台？

Clips Machine 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。

谁开发了 Clips Machine？

由 Mayank8290（@mayank8290）开发并维护，当前版本 v1.2.1。

推荐 Skills
self-improving agent
Captures learnings, errors, and corrections to enable continuous improvement. Use when: (1) A command or operation fails unexpectedly, (2) User corrects Clau...
⬇ 463283 · by pskoett
Skill Vetter
Security-first skill vetting for AI agents. Use before installing any skill from ClawdHub, GitHub, or other sources. Checks for red flags, permission scope, and suspicious patterns.
⬇ 259410 · by spclaudehome
Polymarket
Query Polymarket prediction markets. Check odds, find trending markets, search events, track price movements.
⬇ 232503 · by joelchance
Self-Improving + Proactive Agent
Self-reflection + Self-criticism + Self-learning + Self-organizing memory. Agent evaluates its own work, catches mistakes, and improves permanently. Use when...
⬇ 200423 · by ivangdavila
Github
Interact with GitHub using the `gh` CLI. Use `gh issue`, `gh pr`, `gh run`, and `gh api` for issues, PRs, CI runs, and advanced queries.
⬇ 191113 · by steipete
ontology
Typed knowledge graph for structured agent memory and composable skills. Use when creating/querying entities (Person, Project, Task, Event, Document), linkin...
⬇ 190156 · by oswalpalash

💬 留言讨论