Back to YouTube
Parker Rex DailyJanuary 10, 2026

Automating YouTube Show Notes with AI | Part 1 of Automating Content Series

Learn how to automate YouTube show notes generation using yt-dlp for transcripts, AI for content creation, and GitHub Actions for ongoing automation. Perfect for content creators looking to improve SEO.

Show Notes

Overview

In this video, I walk through automating the creation of show notes for YouTube videos. This is part one of an automating content series where we'll build systems that handle repetitive content tasks without manual intervention.

The key insight: most content creators don't realize that having written versions of their videos creates SEO hooks. Every keyword mentioned becomes searchable content that compounds over time.

The Problem

  • 170+ videos across two channels with no written show notes
  • Manual transcription and formatting is time-consuming
  • Missing out on SEO benefits from searchable text content
  • No system for ongoing videos

The Solution

Backfilling Existing Content

Used a two-step process:

  1. yt-dlp — A library that can download video, audio, or just subtitles/captions from YouTube
  2. AI API — Pass the transcript to an AI with a prompt template based on show notes I liked

The entire script was written by AI. Just explained what I wanted and it built the TypeScript files that handle:

  • Caching logic for API calls
  • Type definitions for video data
  • Functions to check if show notes already exist
  • The actual sync and generate workflow

Ongoing Automation with GitHub Actions

Set up a YAML workflow that runs on a cron schedule:

  • Checks for new videos on both channels
  • Uses yt-dlp to grab transcripts
  • Generates show notes with AI
  • Auto-commits the new MDX files to the repo

The GitHub token enables the action to commit and merge without manual intervention.

Enrichment Pass

First pass didn't include links properly, so I had Claude sweep through all files using sub-agents to:

  • Find show notes missing links
  • Add relevant links where possible
  • Remove low-confidence entries

Key Takeaways

  1. Start with the outcome — Tell AI what you want to achieve, not how to build it
  2. Use examples — Provide a show notes format you like as a template
  3. Backfill first, automate second — Get existing content caught up, then set up ongoing automation
  4. Internal linking — Once you have written content, you can add internal links between related pieces