Edit YouTube Videos Entirely with AI

Editing YouTube videos is the bottleneck nobody talks about. You spend an hour recording. Then you spend four to six hours cutting silences, removing retakes, switching camera angles, adding B-roll, and fixing colors. That's if you're doing it yourself. If you hire an editor, you're looking at $50-200 per video and a 3-5 day turnaround. For weekly uploads, neither option scales. Most creators either burn out on editing or burn cash on editors. And if you're a solopreneur, that editing time comes straight out of the hours you could spend on your actual business. The tools exist to automate this. ffmpeg can do almost anything with video. Remotion can generate professional animations with code. AI can transcribe and understand what you said. But stitching all of that together into a working pipeline? That's the part nobody had figured out. Until now.

This system runs three skills in sequence. Each one handles a different phase of the edit. First, `/video-edit` takes your raw footage and does the grunt work. It syncs your cameras (if you use two angles), transcribes everything with word-level timestamps, identifies retakes by comparing the transcript to your script, removes silences, calibrates face crops for different shot sizes, and generates one massive ffmpeg command that does the entire edit in a single render pass. Then `/video-animate` reads your script, spots the moments that need visual support (data points, process steps, key concepts, quotes worth highlighting), and builds Remotion animations for each one. You get rendered MP4 clips plus a timestamped map showing exactly where each animation goes. Finally, `/video-finalize` takes the edited video and the animation clips and composites everything together. It re-transcribes the edited video (timestamps shift after cuts), matches animations to the right moments, inserts them as overlays or full-frame B-roll, and renders the final output. The whole pipeline runs locally on your machine. No cloud processing, no API costs for the video work. You own everything it produces.

A ready-to-use workflow file that runs on Claude Code. Download it, follow the setup steps, and you're running in minutes.

# Edit YouTube Videos Entirely with AI
 
## What This Skill Does
Record your video. Type one command. Get back a polished video with animations, B-roll, and color correction. This is the exact system behind the viral tweet (100K+ views) where I said 'This video was edited entirely by Claude Code.'
 
## Required Tools
- YouTube
 
## Instructions

 
### Step 1: Load Brand Context
Read the user's CLAUDE.md and /brand folder
to understand target audience and positioning.
 
### Step 2: Research & Analyze
Using available tools, research the market
and competitive landscape for context.
 
### Step 3: Generate Output
Create the deliverables based on the research
and brand context gathered above.

This is a "hard" difficulty idea for a reason. You'll need Claude Code, ffmpeg, Node.js, and faster-whisper installed. The README walks through setup step by step, but expect about 10 minutes to get everything running. The download includes the three Claude Code skills, a complete Remotion project with 6 animation compositions, and example scripts showing what the pipeline actually outputs. The animations use a customizable brand theme, so swap in your colors and fonts to match your channel. Start with a short video (2-3 minutes) for your first run. Get familiar with how the pipeline works before throwing a 15-minute recording at it. And review the segment list after the first skill runs. It's good at catching retakes, but a quick 2-minute scan saves you from re-rendering. The system was built by analyzing how professional YouTube editors work (the kind editing for Alex Hormozi and Matt Gray). It's a starting point that gets better as you customize it for your style.