agentleFS
swyxioswyxio/skills/youtube-thumbnails/SKILL.md

youtube-thumbnails

Skill194 starsChanged 8 months agoMITView on GitHub

This skill should be used when the user asks to "create YouTube thumbnails", "generate thumbnails", "make video thumbnails", or needs AI-generated thumbnail images via Google Gemini for YouTube videos. Covers prompt engineering, image generation, compression, and upload to YouTube Studio.

5 other projects use a version of this, 1 of them edited.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

7 rules. 0 are common, 5 are distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 4 edited versions we read keep these.

  • 5

    Distinctive

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use this prompt template, customized per video:

    line 35
  2. Distinctive:

    Always ask Gemini to search the web for relevant logos and reference images

    line 62
  3. Distinctive:

    Keep the background dark so text pops

    line 66
  4. Distinctive:

    Verify the thumbnail appears in the preview

    line 90
  5. Distinctive:

    After clicking Upload file, use osascript to navigate the native file dialog:

    line 94
  6. Very specific:

    After convert -resize 1280x720 -quality 85: ~200-250KB (well under 2MB limit)

    line 113
  7. Very specific:

    Upload button not visible: Use the find tool to locate the hidden "Upload file" button by ref.

    line 145

“Common” is measured: kept by at least 6 in 10 of the 4 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

YouTube Thumbnail Generation via Gemini

Overview

Generate custom YouTube thumbnails using Google Gemini's image generation (Pro mode). Each thumbnail is a 1280x720 image with bold text overlays, thematic visuals, and branding.

Step-by-Step Workflow

1. Gather Video Info

For each video that needs a thumbnail, collect:

  • Title (already set on YouTube)
  • Content type: "AI in Action" weekly jam or "Paper Club" reading
  • Topic summary: What was presented/demoed (from frame extraction or title)
  • Presenter name (optional, for subtitle text)
  • Date in D Mon YYYY format

2. Navigate to Gemini

  • Go to gemini.google.com
  • Click Tools button (bottom toolbar)
  • Enable Create images
  • Click the model dropdown (shows "Fast" by default) and select Pro
  • Confirm "Image" chip and "Pro" label appear in the chat input area

3. Craft the Thumbnail Prompt

Use this prompt template, customized per video:

Create a YouTube thumbnail (1280x720) for a video titled "[FULL TITLE]".
[1-2 sentences describing what the video is about — the demo, paper, or topic].
The visual should be [THEME DESCRIPTION]: [specific visual elements relevant to the topic].
Bold text overlay: "[MAIN TITLE]" prominently, with smaller text "[SUBTITLE]" below it.
Include a bottom bar with "[SERIES NAME] [D MON YYYY]" in bold.
Please include the Latent Space podcast logo in a corner — search the web for "Latent Space podcast logo" to find it.
Use [COLOR PALETTE] with a dark [THEME] background.

Prompt Customization by Content Type

AI in Action (demos/presentations):

  • Bottom bar text: AI IN ACTION [D MON YYYY]
  • Visual theme: Match the demo topic (space, gaming, dev tools, etc.)
  • Colors: Vibrant, exciting — neon blues, greens, purples, electric tones

Paper Club (paper readings):

  • Bottom bar text: PAPER CLUB [D MON YYYY]
  • Visual theme: Academic but engaging — neural networks, brain visualizations, formulas
  • Colors: Deep blue, orange, white — scholarly but modern
  • Main text: Paper acronym or short name (e.g., "SDPO")
  • Subtitle: Full paper name

Key Prompt Tips

  • Always ask Gemini to search the web for relevant logos and reference images
  • Specify 1280x720 dimensions explicitly
  • Request bold, large text for the main title — readable at thumbnail size
  • Include a colored bottom bar with the series name and date
  • Keep the background dark so text pops

4. Generate and Save

  • Submit the prompt and wait ~30-45 seconds for Pro mode generation
  • Click the generated image to expand it
  • Click Save to download to ~/Downloads
  • Generated files will be named Gemini_Generated_Image_[hash].jpeg

5. Compress to Under 2MB

YouTube requires thumbnails under 2MB. Gemini Pro images are typically 3-4MB.

convert "Gemini_Generated_Image_[hash].jpeg" -resize 1280x720 -quality 85 thumb_[label].jpg

Naming convention: thumb_[abbreviated_date].jpg (e.g., thumb_jan31.jpg, thumb_feb7.jpg)

6. Upload Thumbnails to YouTube Studio

For each video:

  1. Go to YouTube Studio → Content → click on the video
  2. Click the pencil/edit icon (Details tab) in the left sidebar
  3. Scroll to the Thumbnail section
  4. Click the Upload file button (may be hidden — use find tool to locate ref)
  5. Select the correct thumb_[date].jpg file
  6. Verify the thumbnail appears in the preview
  7. Click Save at the top right

File picker automation (macOS): After clicking Upload file, use osascript to navigate the native file dialog:

tell application "System Events"
    delay 1
    keystroke "g" using {command down, shift down}
    delay 1
    keystroke "/path/to/thumb_[date].jpg"
    delay 0.5
    keystroke return
    delay 1
    keystroke return
end tell

7. Verify

Take a screenshot of each video's details page to confirm the custom thumbnail is set.

File Size Reference

  • Gemini Pro raw output: ~3-4MB (too large)
  • After convert -resize 1280x720 -quality 85: ~200-250KB (well under 2MB limit)
  • YouTube max thumbnail size: 2MB
  • YouTube thumbnail dimensions: 1280x720 (16:9)
  • Accepted formats: JPG, PNG, GIF

Example Prompts

Tech Demo Thumbnail

Create a YouTube thumbnail (1280x720) for a video titled "SpaceMolt - AI Agents in Multiplayer Space Games".
This is about a demo of SpaceMolt, a multiplayer space trading and mining game where AI agents autonomously play alongside human players.
The visual should be exciting and space-themed: spaceships, asteroids, neon glow, a dark cosmic background.
Bold text overlay: "SPACEMOLT" prominently, with smaller text "AI AGENTS IN SPACE GAMES" below it.
Include a bottom bar with "AI IN ACTION 7 FEB 2026" in bold.
Please include the Latent Space podcast logo in a corner — search the web for "Latent Space podcast logo".
Use vibrant sci-fi colors — electric blue, green, purple neon tones.

Academic Paper Thumbnail

Create a YouTube thumbnail (1280x720) for a video titled "RL via Self-Distillation (SDPO) Paper Reading".
This is an academic paper reading session about SDPO — Self-Distillation Preference Optimization.
The visual should be academic but engaging: neural network or brain visualization with mathematical formulas.
Bold text overlay: "SDPO" prominently, with smaller text "SELF-DISTILLATION PREFERENCE OPTIMIZATION" below it.
Include a bottom bar with "PAPER CLUB 12 FEB 2026" in bold.
Please include the Latent Space podcast logo in a corner — search the web for "Latent Space podcast logo".
Use deep blue, orange, and white color palette with a dark academic-themed background.

Troubleshooting

  • "File is bigger than 2MB": Compress with ImageMagick convert -resize 1280x720 -quality 85
  • Gemini generation takes long: Pro mode takes 30-45 seconds. Wait and check with a screenshot.
  • Upload button not visible: Use the find tool to locate the hidden "Upload file" button by ref.
  • File picker automation fails: Instruct user to manually select the file.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.