video-editing
AI-assisted video editing workflows for cutting, structuring, and augmenting real footage. Covers the full pipeline from raw capture through FFmpeg, Remotion, ElevenLabs, fal.ai, and final polish in Descript or CapCut. Use when the user wants to edit video, cut footage, create vlogs, or build video content.
Security Assessment
About video-editing
video-editing is a Claude skill that codifies an AI-assisted workflow for cutting, structuring, and augmenting real footage rather than generating video from prompts. Its core thesis is that AI's value in video work is compression: turning long recordings into tight short-form content by transcribing, planning structure, and identifying dead sections, then executing deterministic edits. It activates when a user wants to edit or cut footage, build a vlog, tutorial, or demo, add overlays, subtitles, music, or voiceover, or reframe video for platforms like YouTube, TikTok, and Instagram.
The skill lays out a six-layer pipeline where each tool has one job: capture (Screen Studio or raw camera footage), organization (Claude or Codex to transcribe, plan cuts, and generate an edit decision list), deterministic cuts (FFmpeg commands for extracting segments, batch cutting, concatenating, building proxies, extracting audio, and normalizing loudness), programmable composition (Remotion React/TSX compositions for overlays, motion graphics, and reusable scenes), generated assets (ElevenLabs voiceover and fal.ai music, SFX, and insert visuals via API), and final human polish in Descript or CapCut. It also covers social-media reframing across aspect ratios.
It targets content creators, developers building demo or tutorial videos, and vloggers who are comfortable running command-line tools and calling generation APIs. It suits anyone who wants to speed up the repetitive parts of editing while keeping final creative control, and who already works with FFmpeg, Node/Remotion, and API keys for services like ElevenLabs and fal.ai.
FAQ
What does this skill actually do?
It guides an AI-assisted editing workflow for existing footage — transcribing and planning cuts, generating FFmpeg and Remotion code, calling generation APIs for voiceover/music/inserts, and handing off to Descript or CapCut for final polish. It does not generate whole videos from prompts.
What tools and prerequisites are needed?
The pipeline references Screen Studio, FFmpeg, Node with Remotion, and API access to ElevenLabs and fal.ai (with keys supplied via environment variables), plus Descript or CapCut for final editing. VideoDB is optional for capture and generative audio.
Does it generate video entirely with AI?
No. The stated philosophy is that the value is compression, not generation — AI is used to structure and augment real footage, and generation is limited to assets you actually need such as voiceover, music, thumbnails, or b-roll.
Who is it best suited for?
Creators and developers producing vlogs, tutorials, and product demos who are comfortable with the command line, Remotion/React compositions, and calling third-party media APIs.
Install video-editing
Quick Setup:
- Copy the skill folder to
.claude/skills/ - Claude will automatically detect and use the skill
Repository
affaan-m/ecc