skills.rest
mycelium-hq
mycelium-hqOFFICIAL
Explore all skills by this author →
GitHub

ingest-youtube

Ingest YouTube transcripts into vault Markdown files with metadata.

31|20|Updated Apr 9, 2026
Search Tags:#markdown#subtitles#youtube#yt-dlp#vault#transcripts#ingest
One-click install
npx skills add https://github.com/mycelium-hq/ai-brain-starter --skill ingest-youtube-mycelium-hq
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: ingest-youtube
Source: https://github.com/mycelium-hq/ai-brain-starter/tree/main/skills/ingest-youtube
Command: npx skills add https://github.com/mycelium-hq/ai-brain-starter --skill ingest-youtube-mycelium-hq

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires yt-dlp.

What problem does it solve?

This skill ingests YouTube transcripts into the vault as markdown files, enabling them to be queried by the knowledge graph and decision pipelines while avoiding video downloads.

Core Features & Use Cases

  • Ingest a single video transcript or a channel's recent uploads into the vault as External Inputs/YouTube/<channel>/<YYYY-MM-DD>-<video>.md
  • Idempotent ingestion: re-ingesting the same video overwrites the existing vault file.
  • Extract subtitles from YouTube via yt-dlp (manual/auto subtitles) and gracefully fallback to a stub when none are available; records video metadata for traceability.

Quick Start

Ingest a YouTube video transcript by running the ingest-youtube skill with a video URL to create a vault file.

Frequently Asked Questions about ingest-youtube

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I save YouTube transcripts as markdown files in my vault?▼

To save YouTube transcripts as markdown files, you provide a video URL and the skill extracts subtitles via yt-dlp, writing one file per video under External Inputs/YouTube with frontmatter metadata and transcript text.

Can I ingest subtitles from an entire YouTube channel at once?▼

Yes, you can ingest YouTube channel subtitles by using channel mode to automatically fetch recent uploads, creating individual dated markdown files for each video under the channel's vault directory.

Do I need yt-dlp to extract YouTube video transcripts?▼

Yes, yt-dlp is required to extract YouTube video transcripts, pulling both manual and auto-generated subtitles while gracefully falling back to a metadata stub when subtitles are unavailable.

What happens if I re-ingest the same YouTube video transcript?▼

Re-ingesting the same YouTube video transcript performs an idempotent overwrite, updating the existing markdown file in your vault with the latest subtitle data and metadata without creating duplicates.

Does the skill download the actual YouTube video files?▼

No, the skill avoids video downloads entirely by extracting only YouTube subtitles and metadata, converting transcripts into queryable markdown text for knowledge graphs.

What metadata is included when ingesting YouTube subtitles to markdown?▼

Ingested YouTube subtitle markdown files include frontmatter with video ID, URL, channel, upload date, duration, language, subtitle source, word count, and detected keyword seeds for downstream processing.

Related Skills

Explore More Data & Analytics Skills→
glebisglebis
345

youtube-transcript

Extract YouTube transcripts and metadata into Markdown with YAML frontmatter.

Community
Advanced
ZpankzZpankz
7

hkgb

Bridge CSV or database domain graphs with LLM-extracted entities from PDFs and text in Neo4j.

Community
Advanced
robwestzrobwestz

search

Perform semantic search across MemPalace with knowledge graph cross-referencing.

Community
Intermediate
I-OnlabsI-Onlabs
3

youtube-transcript

Fetch YouTube transcripts via subtitles or Whisper transcription with yt-dlp.

Community
Advanced