How minightdev/autoshorts works
Project type and purpose
• Autoshorts AI is a full-stack web application that ingests long videos (from uploads or third-party sources), runs AI-powered video analysis & editing, and delivers short, shareable clips with effects, captions, and metadata.
• Key use cases: automatic highlight extraction, face tracking, captioning, branded effects, subscription-based usage, cloud storage, and social-media scheduling.
(implied) file structure
• Top level:
• app/ or src/ – contains the Python Flask application code (routes, business logic, AI pipelines)
• templates/ – HTML templates using Bootstrap & jQuery for the UI
• static/ – CSS, JS, images, Font-Awesome, custom styles & scripts
• config/ – Nginx, Gunicorn, Flask settings (environment, secrets, database URIs)
• services/ or workers/ – modules invoking OpenAI Whisper/GPT, Dlib, FFmpeg/MoviePy, OpenCV, ImageMagick, Twitch-dl, Pytube
• tasks/ – background/video-processing task definitions (could be Celery or RQ)
• tests/ – unit and integration tests
• requirements.txt / Pipfile – Python dependencies
• Dockerfile / docker-compose.yml (optional) – deployment definitions
• README.md – project overview
README for additional insights
• Tech stack: Nginx ➔ Gunicorn ➔ Flask, MySQL, Wasabi cloud, AI libraries (Whisper, GPT), video libs, payment SDKs, reCAPTCHA
• Features: authentication, subscription/credits, affiliate system, notifications, scheduling uploads
Elements
• Clear labels for each component (e.g., “Flask API”, “Video Processor Worker”)
• UI components (blue)
• Backend services (green)
• Workers/asynchronous (orange)
• External 3rd-party services (gray)