Skip to content
Back to skills

Get Youtube Transcript Raw

BSecurity

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

  • 666 stars
  • 0 votes
  • 0 copies
  • 2 views
  • Added February 8, 2026
toolspythonshellbashapi

Works with

  • terminal
  • cli
  • api

Security analysis

B75/100
  • criticalModifies startup scripts or system services for persistence

Pro shows the line behind each finding and how to fix it

Scanned February 12, 2026

npx -y skills add majiayu000/claude-skill-registry --skill get-youtube-transcript-raw --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Get Youtube Transcript Raw?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Get Youtube Transcript Raw
[![Security: B — Skills Directory](https://www.skillsdirectory.com/api/skills/majiayu000-get-youtube-transcript-raw/badge)](https://www.skillsdirectory.com/skills/majiayu000-get-youtube-transcript-raw)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: get-youtube-transcript-raw
description: Capture a YouTube video transcript as raw material using `ytt`, storing it in the raw/ directory with minimal metadata for later distillation.
---

# 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=...` or `https://youtu.be/...`)

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

## Outputs

This skill produces:
1. A new Markdown file in `raw/` named `YYYYMMDD-HHMMSSZ--<slug>.md`
2. 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)
3. 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):

```bash
./scripts/ytraw "<youtube_url>"
```

If the current environment can’t access YouTube (common in sandboxes), run `ytt` locally and pipe:

```bash
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:**
1. Copy the YouTube URL to your clipboard.
2. Run the following command in your terminal:
   ```bash
   ./scripts/ytraw
   ```
3. 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`:

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

## References

- Pipeline and front matter schema: `docs/distillation/distillation-pipeline.md`

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.