Claude-skill-registry get-youtube-transcript-raw

Capture a YouTube video transcript as raw material using `ytt`, storing it in the raw/ directory with minimal metadata for later distillation.

install

source · Clone the upstream repo

git clone https://github.com/majiayu000/claude-skill-registry

Claude Code · Install into ~/.claude/skills/

T=$(mktemp -d) && git clone --depth=1 https://github.com/majiayu000/claude-skill-registry "$T" && mkdir -p ~/.claude/skills && cp -r "$T/skills/data/get-youtube-transcript-raw" ~/.claude/skills/majiayu000-claude-skill-registry-get-youtube-transcript-raw && rm -rf "$T"

manifest: skills/data/get-youtube-transcript-raw/SKILL.md

source content

Get YouTube Transcript (Raw)

When to use

Use when you need the raw transcript (plus YouTube title/description) saved to

raw/

for later distillation.

Keywords: youtube, transcript, captions, ytt, raw, capture

Inputs

Required:

url

(string): YouTube URL (e.g.,

https://www.youtube.com/watch?v=...

https://youtu.be/...

)

Optional:

```
title_hint
```
(string): Used only if
```
ytt
```
can’t provide a title.

Outputs

This skill produces:

A new Markdown file in
```
raw/
```
named
```
YYYYMMDD-HHMMSSZ--<slug>.md
```

YAML front matter aligned with

docs/distillation/distillation-pipeline.md

```
title
```
(best-effort)
```
source_url
```
(the provided URL)
```
captured_at
```
(UTC ISO timestamp)
```
capture_type: youtube_transcript
```
```
capture_tool: ytt
```
```
raw_format: markdown
```
```
status: captured
```
(or
```
capture_failed
```
on failure)

Body content: the raw transcript text emitted by
```
ytt
```
(no summarization).

Prerequisites

```
ytt
```
(this repo’s YouTube transcript utility)
```
python3
```
(used by the bundled script for slugging and safe YAML string escaping)

Quick start

Capture directly (uses

ytt fetch --no-copy

internally):

./scripts/ytraw "<youtube_url>"

If the current environment can’t access YouTube (common in sandboxes), run

ytt

locally and pipe:

ytt fetch --no-copy "<youtube_url>" | ./scripts/ytraw

Manual execution (Fallback)

If you encounter persistent issues capturing a transcript within the sandbox (e.g., network restrictions or tool failures), inform the user they can run the script manually on their local machine.

The script

./scripts/ytraw

is designed to extract the URL directly from the clipboard if no arguments are provided.

Instructions for the user:

Copy the YouTube URL to your clipboard.
Run the following command in your terminal:
```
./scripts/ytraw
```
If manual adjustments are required, you can edit the file directly. If you prefer the model to perform adjustments, share the file path or URL with it.

Optional: shell alias (zsh)

Add to

~/.zshrc

ytraw () { ytt fetch --no-copy "$1" | /path/to/repo/scripts/ytraw --from-stdin }

References

Pipeline and front matter schema:

docs/distillation/distillation-pipeline.md