Skip to main content
skill CX Codex

youtube-transcript · glebis — Claude Code skill: extract YouTube subtitles

The youtube-transcript skill pulls subtitles and video metadata from YouTube, saves them as Markdown with YAML frontmatter into an Obsidian vault.

MIT 258 installs ★ 414 updated 2026-10-23

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓
npx skills add glebis/claude-skills --skill youtube-transcript -a claude-code
Open on GitHub

What youtube-transcript does

The skill runs yt-dlp to fetch video metadata (title, channel, upload date, tags, view count) and subtitles in VTT format. Subtitles are grouped by video chapters if present, with timestamps for each segment. The file is saved to ~/Brains/brain/ with a sanitized filename based on the video title.

The script looks for English subtitles first, falls back to Russian if English is unavailable. Video and audio files are not downloaded. The description is truncated to 500 characters. Requires yt-dlp installed (pip install yt-dlp). A bundled deduplicate_transcript.py script removes duplicate lines from processed files — VTT format sometimes accumulates repeated entries.

The skill works with Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, and OpenCode.

What it helps with

  • extract subtitle text from YouTube videos
  • save video metadata in YAML format
  • split transcript by video chapters
  • remove duplicate lines from VTT subtitles

Example prompts for the agent

  • Download subtitles from this video and save to Obsidian
  • Extract the transcript from this YouTube link and split by chapters
  • Run extract_transcript.py for this YouTube URL

How to install youtube-transcript

1Skills CLI (Vercel)

npx skills add glebis/claude-skills --skill youtube-transcript -a claude-code

2Manually: put the files in place

~/.claude/skills/youtube-transcript/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/youtube-transcript/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about youtube-transcript

How to install youtube-transcript in Claude Code?

Run npx skills add glebis/claude-skills --skill youtube-transcript -a claude-code, the skill will appear in ~/.claude/skills/youtube-transcript/SKILL.md.

What does youtube-transcript do?

The skill extracts subtitles and metadata from YouTube videos using yt-dlp, formats them as Markdown with YAML frontmatter, and saves them to an Obsidian vault. Video and audio files are not downloaded.

What does youtube-transcript require?

Requires yt-dlp installed (pip install yt-dlp). The video must have subtitles available — auto-generated or manual, in English or Russian.

Source text

SKILL.md · MIT
---
name: youtube-transcript
description: "Extract YouTube video transcripts with metadata and save as Markdown to Obsidian vault. Use this skill when the user requests downloading YouTube transcripts, converting YouTube videos to text, or extracting video subtitles. Does not download video/audio files, only metadata and subtitles."
---

# YouTube Transcript

## Overview

Extract YouTube video transcripts, metadata, and chapters using yt-dlp. Output formatted as Markdown with YAML frontmatter, saved to ~/Brains/brain/ (Obsidian vault).

## Quick Start

To extract a transcript from a YouTube video:

```bash
python scripts/extract_transcript.py <youtube_url>
```

Optional: Specify custom output filename:

```bash
python scripts/extract_transcript.py <youtube_url> custom_filename.md
```

## Output Format

### YAML Frontmatter

The generated Markdown includes comprehensive metadata:

- `title` - Video title
- `channel` - Channel name
- `url` - YouTube URL
- `upload_date` - Upload date (YYYY-MM-DD)
- `duration` - Video duration (HH:MM:SS)
- `description` - Video description (truncated to 500 chars)
- `tags` - Array of video tags
- `view_count` - View count
- `like_count` - Like count

### Body Structure

Transcript organized by video chapters (if available):

```markdown
## Chapter Title

**00:05:23** Transcript text for this segment.

**00:05:45** Next segment text.
```

If no chapters exist, all content appears under "## Transcript" heading.

Timestamps formatted as HH:MM:SS for consistency.

## Workflow

1. Extract metadata and subtitles using yt-dlp
2. Parse VTT subtitle format to extract timestamps and text
3. Group transcript segments by video chapters (if present)
4. Format as Markdown with YAML frontmatter
5. Save to ~/Brains/brain/ with sanitized filename based on video title
6. Clean up temporary subtitle files

## Deduplication

To remove duplicates from existing transcript files:

```bash
python scripts/deduplicate_transcript.py <markdown_file>
```

This removes transcript entries that are prefixes of subsequent entries (common in VTT files where subtitles accumulate).

## Requirements

Ensure yt-dlp is installed:

```bash
pip install yt-dlp
```

## Limitations

- Extracts subtitles in English first, falls back to Russian if English unavailable
- Requires video to have subtitles (auto-generated or manual)
- Does not download video or audio files
- Description truncated to 500 characters in frontmatter

© glebis · MIT · glebis/claude-skills

Similar skills

All skills for Notes, tasks and projects →

youtube-transcript by other authors comes first.

youtube-transcript badlogic youtube-transcript — a skill that pulls a YouTube video's caption track and delivers it line by line with timestamps. It accepts a bare video ID or any of the three common YouTube URL formats. CX Codex +1 930installs youtube-transcript intellectronica youtube-transcript extracts transcripts and subtitles from YouTube videos. Useful when you need video text with or without timestamps inside your agent. CX Codex +1 3,500installs youtube-transcript michalparkola youtube-transcript is a Claude Code skill that downloads YouTube subtitles or turns a video’s audio into a transcript. It activates for a video link or a request to fetch captions or transcribe a video. CX Codex +1 908installs youtube-transcript tubeagentkit Pulls YouTube transcripts and channel metadata. Triggers on requests to summarize, quote, or search spoken video content. CX Codex +1 ★ 41repo stars baoyu-youtube-transcript JimLiu baoyu-youtube-transcript pulls YouTube transcripts and cover images from a URL or video ID. Handy when you want the text of a video without copying it by hand. CX Codex +1 23,100installs youtube-fetcher JimmySadek youtube-fetcher — a skill for Claude Code and five other agent environments pulls captions and transcribes video speech. A bare link saves an Obsidian note; a question about the video gets answered without writing files. CX Codex +1 ★ 546repo stars YouTube Transcript youtube YouTube Transcript pulls the caption text of a YouTube video and hands it to your agent as timestamped segments, so a summary or a question about the content is one step away. CX Codex +1 543installs mcp-youtube-transcript jkawamoto mcp-youtube-transcript — an MCP server that retrieves YouTube video transcripts, timed segments, and metadata for AI agents. It splits long transcripts into chunks to protect the LLM context window. CX Codex +1 ★ 499repo stars youtube Eugeniu Ghelbur Pulls a YouTube transcript, metadata, and top comments into an Obsidian note. Add --visual when the video’s frames matter too. CX Codex +1 ★ 5,412repo stars youtube-transcribe-skill feiskyer youtube-transcribe-skill pulls captions from a YouTube video and writes them to a timestamped text file. Handy when you need a transcript in your project directory. CX Codex +1 366installs

From the same set51

The whole set →
telegram Fetches, searches and sends Telegram messages, downloads attachments, edits sent messages. Works via telegram_dl and Telegram API. CX Codex +1 287installs brand-agency brand-agency pushes Agency's neobrutalist look into slides, SVG, HTML and social images: fixed palette, font stack, hard-shadow rules. CX Codex +1 276installs doctorg doctorg is a Claude Code skill that answers health, nutrition, training and sleep questions by checking trusted sources and grading the strength of evidence for each claim. CX Codex +1 265installs tdd tdd runs test-driven development cycles with strict context separation across subagents. Useful for writing clean features and reproducing bugs. CX Codex +1 262installs gmail Skill for searching Gmail messages via API with filters for sender, date, status and attachments. Works on Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI and OpenCode. CX Codex +1 258installs transcript-analyzer transcript-analyzer extracts decisions, action items, opinions, unresolved questions, and domain terms from meeting transcripts using Cerebras AI. CX Codex +1 258installs zoom Skill to create, update, and list Zoom meetings, and access cloud recordings via Zoom API. Search recordings with transcripts, summaries, and download links. CX Codex +1 252installs fathom Skill to retrieve meetings, transcripts, AI summaries, and action items directly from Fathom API. Use it to sync meeting recordings and transcripts into your coding workflow. CX Codex +1 249installs

More in the catalog

Project sponsors

Companies that keep this analytics open