ingest-youtube

What problem does it solve?

Manually retrieving, cleaning, and integrating YouTube transcripts into a knowledge vault is slow and error-prone; this Skill automates subtitle extraction, cleaning, and saving so video content can be reliably ingested into a wiki.

Core Features & Use Cases

• Automatic subtitle extraction: uses yt-dlp to download auto-generated or manual English subtitles and extract video metadata.
• SRT cleaning and deduplication: removes sequence numbers, timestamps, and duplicate lines to produce readable transcript text.
• Lazy dependency installation: will attempt to install yt-dlp via Homebrew on first use if missing.
• Vault integration: saves cleaned transcripts to the vault's raw/ directory with YAML frontmatter including source-url, title, author, and date.
• Use Case: Ingest a lecture or conference talk to create a searchable wiki page with transcript and metadata for research or documentation.

Quick Start

Run the ingest-youtube skill with the video URL to fetch metadata, download English subtitles, clean the SRT, and save a transcript ready for wiki ingestion.