YouTube SEO Pack: Video → Titles, Description, Chapters, Tags & 3 Thumbnails
Point your agent at your video file and it writes **and designs** everything YouTube Studio asks for, based on what you actually said in the video.
$12
Works with the AI Agents you already use
YouTube SEO Pack: Video → Titles, Description, Chapters, Tags & 3 Thumbnails
Example session with this skill installed
Here's my video, do the full upload kit with 3 thumbnails.
- Read your context and instructions
- Compiled the youtube seo
youtube-seo-pack.md with 10 ranked titles (⭐ Buying DJI Pocket 4 in Bangkok: 3 Stores, 15,500 Baht), a paste-ready description, 12 chapters starting at 0:00, tags at 470/500 characters, hashtags, a pinned comment and a validation checklist; plus description.txt, tags.txt, chapters.txt and thumbnail_A/B/C.jpg with preview.jpg.
Connects securely to your tools. The creator never sees your data.
About this skill
Stop staring at the upload screen. Point your agent at your video file and it writes and designs everything YouTube Studio asks for, based on what you actually said in the video.
From your video:
- Transcribes locally with ffmpeg + Whisper. Nothing is uploaded. It uses your NVIDIA GPU automatically (a 14-minute vlog took 53 seconds in testing) and falls back to CPU if there isn't one.
- Warns you when the language guess is unsure. Mixed-language speech is the #1 cause of garbage transcripts.
- Splits captions into readable cues of 7 seconds or less, using word-level timing.
The SEO pack:
- 10 titles across proven angles, ranked, with character counts
- A description with your keyword and hook in the first 150 characters
- Chapters that YouTube accepts: they start at 0:00, there are at least 3, and each is at least 10 seconds long
- Tags that fit the 500-character limit, ordered by importance
- 3-5 hashtags and a pinned comment that asks a real question
- A validation checklist, so nothing gets cut off or rejected
3 thumbnails (1280x720, under 2 MB, ready for YouTube Test & Compare):
- Scans your footage, or your raw clips folder, for the sharpest, eyes-open poses and clean, colourful background scenes, and shows them on contact sheets
- Uses your own photo instead if you have one
- Removes the background and adds an outline and drop shadow, sizing you by your face so every thumbnail has a big, readable head
- Uses a scene background for vlogs and a plain or gradient one with a glow for tutorials and talking heads
- Auto-fits the bold text, with an optional price badge and arrow
- Previews everything at feed size and warns when the text is too small to read
Works with .mp4/.mov/.mkv, .srt, .vtt, Whisper JSON and plain text; long-form and Shorts; any language; one video or a whole folder.
Honest by design: it never invents prices, facts or links. It asks before installing anything, and never overwrites files it didn't create.
Requirements:
- Text pack from a transcript: nothing.
- From a video: Python 3.9+ and ffmpeg.
- Thumbnails: opencv, pillow and rembg.
The included setup_check.py tells you exactly what's missing and gives the install commands. It installs nothing itself. One-time model downloads: Whisper (0.5-1.6 GB) and the background-removal model (about 170 MB).
Example output (from a real test: a 14-minute travel vlog in Urdu)
⭐ Top title: Buying DJI Pocket 4 in Bangkok: 3 Stores, 15,500 Baht
Chapters (12): 0:00 Bought the DJI Pocket 4 in Bangkok · 0:56 Grab ride to the DJI store · 2:50 Central Mall: the wrong floor · 4:49 DJI store 1: only the Creator Combo · … · 12:48 Cheaper than Pakistan? Customs and VAT refund
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse All Claude Skills.
Keywords are only marked verified if your agent has web search. Doesn't upload to YouTube or invent links (missing links become [LINK: ...] placeholders). Chapters need a timestamped transcript. Default thumbnail font covers Latin text only; pass a font for Urdu/Arabic/Hindi.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed today
- Software
Python 3.9+ (video/thumbnails)ffmpeg (video input)
- Optional
- faster-whisperopencv-python, pillow, numpyrembg
Text pack from a transcript needs nothing. setup_check.py lists what's missing and the install commands; the agent asks before installing. One-time model downloads: Whisper 0.5-1.6 GB, background removal ~170 MB.
Permissions
Allowed Hosts
File Scopes
Scripts make no network calls of their own; video, audio and images never leave your computer. Asks before installing anything.