Automate talking-head video editing with a 3-pass cut pipeline and subtitles.
npx skills add https://github.com/cjrfoto/claude-config --skill yt-cut-edit
Please help me install this Agent Skill. Skill: yt-cut-edit Source: https://github.com/cjrfoto/claude-config/tree/main/skills/yt-cut-edit Command: npx skills add https://github.com/cjrfoto/claude-config --skill yt-cut-edit
This Skill eliminates the tedious manual work of editing talking-head videos by performing a deterministic 3-pass pipeline that removes retake blocks, re-transcribes for detail cuts, and polishes silences to deliver sentence-aware subtitles.
Provide a video folder path and optional --speed and --script arguments to output a fully edited cut with subtitles.
High-intent search queries and answers about installing and using this skill.
Automate talking-head video editing by running a 3-pass cut pipeline that removes retake blocks, re-transcribes for detail cuts, and polishes silences to output a fully edited video with sentence-aware subtitles.
A 3-pass cut pipeline works by first removing retake blocks, then re-transcribing the remaining footage for detail cuts, and finally applying silence polish with sentence-aware subtitles while preserving 100% of scripted content.
Yes, you need ffmpeg for video rendering, a Deepgram API key for transcription, Python, and Claude Code for decision logic to execute the automated editing and subtitling pipeline.
Yes, you can apply a single global speed adjustment after all cuts are completed and before subtitling by passing the optional --speed argument when providing the video folder path.
The best way to ensure script fidelity is using a deterministic workflow where Claude Code handles decision logic for cut zone generation, preserving 100% of scripted content plus roughly 2% improvisation.
Extract transcripts and generate cut zones by providing a video folder path and optional --script argument to Claude Code, which processes the original footage and renders outputs into assets/edit.
yizhiyanhua-aiSearch and download images and video clips from Pexels, Pixabay, Unsplash, and YouTube.
antoniolgTranscribe videos with Whisper, burn subtitles, and schedule posts via Postiz MCP.
clsandovalTranscribe WhatsApp voice notes to text using OpenAI Whisper.
different-aiTranscribe video speech, detect filler words and pauses, and generate FFmpeg cut commands with SRT subtitles.
