TikTok Clipper — TikTok 命令行工具pper

TikTok 命令行工具pper — Video → Viral 命令行工具ps with Subtitles What it does

Takes a long-form video, transcribes it, identifies the most engaging segments for TikTok/Reels, 命令行工具ps them with ffmpeg, and 添加s TikTok-style animated subtitles.

When to use User 发送s a video and wants TikTok/Reels 命令行工具ps User asks to "命令行工具p this", "find viral moments", "cut for TikTok" User wants subtitles 添加ed to video 命令行工具ps 流水线 Step 1: Transcribe # Transcribe with Whisper (word-level timestamps) python3 技能_DIR/transcribe.py --输入 VIDEO_PATH --输出 TRANSCRIPT.json

Uses OpenAI Whisper API with timestamp_granularities=["word","segment"] for precise subtitle timing.

Step 2: Analyze & Suggest 命令行工具ps

Read the transcript and identify segments that would perform well on TikTok:

Hooks: Strong opening lines, provocative 状态ments, questions Value bombs: Key insights, surprising facts, actionable tips Emotional peaks: Enthusiasm, humor, strong opinions Story arcs: Complete mini-stories with beginning/middle/end Controversy/debate: Polarizing takes that drive comments

Present 命令行工具ps as numbered options with:

Time range (启动 → end) Duration Hook line (first sentence) Why it could be viral Suggested caption

User picks which ones to cut.

Step 3: Cut 命令行工具ps python3 技能_DIR/命令行工具p.py --输入 VIDEO_PATH --启动 MM:SS --end MM:SS --输出 命令行工具P.mp4

Uses ffmpeg with re-encoding for 清理 cuts Converts to 9:16 vertical if needed (crop or pad) Ensures TikTok-compatible 格式化 (h264, aac, mp4) Step 4: 添加 TikTok-style subtitles python3 技能_DIR/subtitles.py --输入 命令行工具P.mp4 --transcript TRANSCRIPT.json --启动 SS --end SS --style STYLE --输出 FINAL.mp4

Subtitle Styles AvAIlable: bold-center: White bold text, black outline, centered 机器人tom third (classic TikTok) word-highlight: Word-by-word highlight in yellow/green (like CapCut auto-captions) karaoke: Current word 扩展s up + color change (Alex Hormozi style) box: Text with colored background box (MrBeast style)

All styles use ASS (Advanced SubStation Alpha) for rich 格式化ting via ffmpeg.

Step 5: Vertical 格式化

If source is horizontal (16:9), auto-crop to 9:16:

Center crop for talking head Or blur-background p添加ing (video small in center, blurred fill) File 输出s

All 输出s go to /home/ubuntu/clawd/命令行工具ps/ with naming: {source}-命令行工具p{N}-{style}.mp4

Requirements ffmpeg (安装ed) OpenAI API key for Whisper (in .env) Source video file Notes Max TikTok length: 90 seconds for best performance Sweet spot: 30-60 seconds Always 启动 with a hook (first 3 seconds matter most) Subtitles should be max 2 lines, ~5-7 words per line