How To use GitSkit.

Everything you need to get from zero to your first supercut, step by step.

Setup

01

Clone and install

Clone the GitSkit repo and install dependencies.

git clone https://github.com/yusufsafary/gitskit
cd gitskit
npm install
02

Configure environment

Copy the example env file and fill in your values.

cp .env.example .env

Edit .env with your GITSKIT_SECRET and optionally your XAI_API_KEY for transcription.

03

Add your videos

Create the bucket structure and drop in your media files.

videos/
  my-podcast/
    videos/     <- put .mp4 / .mp3 files here
    transcripts/
    supercuts/
04

Run GitSkit

Start the dev server and open it in your browser.

npm run dev

Environment variables

GITSKIT_SECRET Yes
Random 32+ char secret for signing sessions
GITSKIT_PASSWORD No
Password to protect access (empty = open)
XAI_API_KEY For transcription
API key from x.ai
SYNTAX_BUCKETS_DIR No
Absolute path to your videos root

Transcription

Once your videos are in place, head to the Transcribe page and queue up your files. GitSkit sends audio to xAI's speech-to-text API and saves word-level JSON transcripts alongside your media.

Transcription is one-time per file. Re-transcribing overwrites the existing transcript.

Building supercuts

Go to the Supercut page. Enter a word, phrase, or regex pattern. GitSkit searches all transcripts, slices the matching clips from the source video using FFmpeg, and stitches them into a single output file.

The rendered clip appears in your Clips library and can be downloaded or deleted.

Deploy to Vercel

GitSkit is designed to deploy to Vercel. Set your environment variables in the Vercel dashboard and push your repo.

vercel --prod
Deploy on Vercel See a demo