youtube-transcribe

youtube-transcribe is a skill for Claude Code, Codex from kennyzir/7deer_skills. It costs 83 tokens per session (2,040 once invoked), scanned A, original, MIT.

A workflow for downloading a YouTube video's audio, converting it, transcribing it with Whisper, and saving the resulting content to memory.

In plain words
What is it for?
Use it when given a YouTube URL and asked for a transcript or content extraction. It relies on yt-dlp, ffmpeg, and the Whisper command-line tool.
Why use it?
It removes the manual work of extracting spoken content from videos and turning it into searchable notes.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one.

Good fit Use it when given a YouTube URL and asked for a transcript or content extraction. It relies on yt-dlp, ffmpeg, and the Whisper command-line tool.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/kennyzir/7deer_skills/youtube-transcribe
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent
npx skills add kennyzir/7deer_skills --skill youtube-transcribe
Clone the repo
git clone --depth 1 https://github.com/kennyzir/7deer_skills

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for youtube-transcribe

README.md
[![agentmods](https://agentmods.dev/badge/skills/kennyzir/7deer_skills/youtube-transcribe/github.svg)](https://agentmods.dev/skills/kennyzir/7deer_skills/youtube-transcribe)
Your own site
<a href="https://agentmods.dev/skills/kennyzir/7deer_skills/youtube-transcribe"><img src="https://agentmods.dev/badge/skills/kennyzir/7deer_skills/youtube-transcribe/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for youtube-transcribe

Your own site · 80×15
<a href="https://agentmods.dev/skills/kennyzir/7deer_skills/youtube-transcribe"><img src="https://agentmods.dev/badge/skills/kennyzir/7deer_skills/youtube-transcribe.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 83 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 2,040 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 1 finding. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector warn 7 Sept 2026
SkillSpector: 3 findings, up to high

These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →

  • high Tool Misuse · line 96
    Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).
    Fix: Validate all tool parameters against an allowlist. Reject dangerous parameter values (shell=True, --force, -rf /) and use safe defaults.
  • high Tool Misuse · line 158
    Tool parameters are crafted to achieve unintended or unsafe behavior. Parameter abuse can bypass intended safety checks (e.g. shell=True, --force, dangerous glob patterns).
    Fix: Validate all tool parameters against an allowlist. Reject dangerous parameter values (shell=True, --force, -rf /) and use safe defaults.
  • medium Rogue Agent · line 181
    Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.
    Fix: Remove any persistence mechanisms (cron jobs, startup scripts, state files). Skills should not maintain state across sessions without explicit user consent.
How audits are shown
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00083 $0.02040
Opus 5 $0.00042 $0.01020
Sonnet 5 $0.00017 $0.00408
Haiku 4.5 $0.00008 $0.00204

Measured 13d ago against content hash cb912db043dc, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-12, from the pricing page.

Security

Grade A, and why

youtube-transcribe scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 13d ago.

The scan reads SKILL.md. This mod also ships 1 executable file (scripts/transcribe.sh), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

curl -sL "https://github.com/conda-forge/miniforge/releases/latest/download/Miniforge3-MacOSX-arm64.sh" -o /tmp/miniforge.sh
youtube-transcribe/SKILL.md · 184 lines

How it starts

The opening of the file, as written. The whole thing — 184 lines — stays where its author put it; the contents beside it link to each section on GitHub.

YouTube Transcribe Skill

Tool Discovery

Before running, the agent checks for available tools and sets PATH:

# Find tools dynamically — don't hardcode paths
export PATH="/tmp/miniforge/bin:$(python3 -m site --user-base)/bin:$PATH"

YTDLP=$(command -v yt-dlp 2>/dev/null || echo "yt-dlp")
FFMPEG=$(command -v ffmpeg 2>/dev/null || echo "ffmpeg")
WHISPER=$(command -v whisper 2>/dev/null || echo "whisper")

# Verify tools exist
for TOOL in "$YTDLP" "$FFMPEG" "$WHISPER"; do
  [ -x "$TOOL" ] || echo "[WARN] Tool not found or not executable: $TOOL"
done

Tool requirements:

Tool Install Fallback path
yt-dlp pip3 install yt-dlp $(python3 -m site --user-base)/bin/yt-dlp
ffmpeg conda install -c conda-forge ffmpeg /tmp/miniforge/bin/ffmpeg
whisper pip3 install openai-whisper $(python3 -m site --user-base)/bin/whisper

Environment PATH

export PATH="/tmp/miniforge/bin:$(python3 -m site --user-base)/bin:$PATH"

Workflow

Step 1 — Parse YouTube URL

URL="https://www.youtube.com/watch?v=Q5kYrmzNhcU"
VIDEO_ID=$(echo "$URL" | grep -oE 'v=[^&]+' | cut -d= -f2 | head -1)
# Handles: https://youtu.be/ID, https://www.youtube.com/watch?v=ID&t=..., https://youtube.com/embed/ID

Step 2 — Get Video Metadata

TITLE=$($YTDLP --extractor-args "youtube:player_client=android" \
  --print title --no-warnings "https://www.youtube.com/watch?v=${VIDEO_ID}" 2>/dev/null)
CHANNEL=$($YTDLP --extractor-args "youtube:player_client=android" \
  --print channel --no-warnings "https://www.youtube.com/watch?v=${VIDEO_ID}" 2>/dev/null)
DURATION=$($YTDLP --extractor-args "youtube:player_client=android" \
  --print duration_string --no-warnings "https://www.youtube.com/watch?v=${VIDEO_ID}" 2>/dev/null)

Step 3 — Download Audio (with Fallback Chain)

mkdir -p /tmp/yt_audio

# Strategy: try android client first → if GVS PO Token error, fall back to web client
# (web client may 403 on some videos; android client needs PO token for high-quality formats
# but usually succeeds with format 18 even without PO token)

# Attempt 1: android client (works without PO token for format 18)
$YTDLP -x --audio-format mp3 --audio-quality 0 \
  --extractor-args "youtube:player_client=android" \
  -f "best[ext=mp4]/best" \
  -o "/tmp/yt_audio/${VIDEO_ID}.%(ext)s" \
  "https://www.youtube.com/watch?v=${VIDEO_ID}" 2>&1 | grep -v "^Deprecated\|^NotOpenSSL\|^Warning:"

# If android fails (GVS PO Token required), fall back to web
if [ ! -f "/tmp/yt_audio/${VIDEO_ID}.mp4" ] && [ ! -f "/tmp/yt_audio/${VIDEO_ID}.mp3" ]; then
  echo "[*] Android client failed, trying web client..."
  $YTDLP -x --audio-format mp3 --audio-quality 0 \
    -o "/tmp/yt_audio/${VIDEO_ID}.%(ext)s" \
    "https://www.youtube.com/watch?v=${VIDEO_ID}" 2>&1 | grep -v "^Deprecated\|^NotOpenSSL"
fi

Read the full file on GitHub · 184 lines

Files

What ships with it

2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 13d ago First seen · 184 lines · 83 tokens per session scan A cb912db043dc

Subscribe to this mod's changes

youtube-transcribe is a skill published in the GitHub repository kennyzir/7deer_skills (313 stars, last pushed 4d ago), licensed MIT. It adds 83 tokens to every session and 2,040 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other skills, from other repositories