A fully autonomous, serverless content generation pipeline engineered to conceptualize, stitch, and publish YouTube Shorts continuously without any human intervention. This project acts as a completely independent media producer, converting idle cloud compute time into passive social media growth.

Built entirely in Python, the system is orchestrated via GitHub Actions cron jobs that spin up temporary cloud runners daily. The pipeline begins by fetching trending topics or reading from a prompt queue. It then interfaces with the OpenAI API to craft highly engaging, retention-optimized short scripts, alongside auto-generating SEO metadata (titles, tags, descriptions). It synthesizes natural-sounding voiceovers using Text-to-Speech APIs, and programmatically edits the video using MoviePy and FFmpeg. This involves downloading satisfying background clips (like kinetic sand or ASMR), overlaying the generated audio, rendering dynamic captions word-by-word for high engagement, and finally authenticating with the YouTube Data API v3 to upload and publish the Short directly to the channel.