This workflow runs on a schedule you set (default is every 10 minutes). It fetches all videos from a specific YouTube playlist, retrieves details and transcripts for each video using RapidAPI, cleans the transcript text, and then uses OpenAI (GPT models) to generate detailed notes, YAML frontmatter, and internal links formatted for Obsidian. Finally, it assembles everything into a Markdown file, saves it to your designated Dropbox folder, and removes the processed video from the source YouTube playlist.
Getting started should take about 5 minutes. A Quick Start Guide (PDF) is included in the download with step-by-step instructions!
Here's a quick overview of the setup:
yt-api
service. The template uses ={{ $env["RAPIDAPI_API_KEY"] }}
. You can either set up an environment variable in n8n named RAPIDAPI_API_KEY
or simply delete that expression in the "Get Video Details" and "Get Video Transcript" nodes and paste your API key directly into the x-rapidapi-key
field.Bonus Trick: Symlink the Dropbox folder (using the Dropbox desktop app) to a folder in your Obsidian vault for beautifully organized YouTube summary notes!