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.
npx skills add deeployCO/youtube-seo-skills --skill youtube-seo-thumbnailgit clone --depth 1 https://github.com/deeployCO/youtube-seo-skillsWrote 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.
[](https://agentmods.dev/skills/deeployco/youtube-seo-skills/youtube-seo-thumbnail)<a href="https://agentmods.dev/skills/deeployco/youtube-seo-skills/youtube-seo-thumbnail"><img src="https://agentmods.dev/badge/skills/deeployco/youtube-seo-skills/youtube-seo-thumbnail/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.
<a href="https://agentmods.dev/skills/deeployco/youtube-seo-skills/youtube-seo-thumbnail"><img src="https://agentmods.dev/badge/skills/deeployco/youtube-seo-skills/youtube-seo-thumbnail.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00065 | $0.01851 |
| Opus 5 | $0.00032 | $0.00925 |
| Sonnet 5 | $0.00013 | $0.00370 |
| Haiku 4.5 | $0.00006 | $0.00185 |
Grade A, and why
youtube-seo-thumbnail scanned grade A with 0 findings 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 12d ago.
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.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 213 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Thumbnail Analysis (Advanced)
Diagnose CTR potential quantitatively when possible, qualitatively when not. Thumbnail is the single biggest pre-click lever after title.
Data Collection
- Source thumbnail: download the maxres (
maxresdefault.jpg) fromhttps://i.ytimg.com/vi/{VIDEO_ID}/maxresdefault.jpg, fall back tohqdefault.jpgif missing - SERP grid: fetch the top 12 YouTube results for the primary keyword, download each thumbnail, build a 4x3 contact sheet
- Computer-vision pass via
scripts/analyze_thumbnail.py:- OpenCV face detection (count, bbox, size %)
- MediaPipe face-mesh for emotion inference (or DeepFace fallback)
- Color histogram (HSV) — dominant 3 colors
- Contrast score (stddev of luminance)
- Text region detection (EAST or Tesseract) — estimated word count and coverage %
- CLIP embedding (
ViT-B/32) for similarity scoring vs SERP grid
- Mobile preview: resize to 120x68, 246x138, 480x270 and inspect legibility manually
Analysis Dimensions
1. Technical
- Resolution 1280x720, 16:9, max 2MB
- Format JPG / PNG
- Safe area: no critical content in bottom-right (duration overlay ~6% of area) or top-right (menu dots)
2. Legibility at small sizes (binding constraint)
At 120x68:
- Subject identifiable? (pass/fail)
- Text readable? (pass/fail)
- Emotion readable? (pass/fail)
If any fail → critical issue. Most thumbnails designed on a desktop monitor at 500px+ fail at 120px.
3. Composition — Gestalt principles
- Figure-ground separation: subject clearly separated from background (rim light, cutout, color contrast, background blur)
- Closure: viewer's brain completes an implied shape or story — works against scroll blindness
- Continuity: eye travels in a designed path (eye direction of face, arrow of object, gradient)
- Similarity/proximity: elements that belong together are grouped
- Rule of thirds: focal point at intersections, not dead-center
- Depth: foreground / midground / background create dimensionality
- Negative space: enough breathing room for the focal point to pop at small sizes
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.
- 12d ago First seen · 213 lines · 65 tokens per session scan A 16edf3b140e0
youtube-seo-thumbnail is a skill published in the GitHub repository deeployCO/youtube-seo-skills (9 stars, last pushed 5mo ago), licensed MIT. It adds 65 tokens to every session and 1,851 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
cms-metadata-writer
Writes structured metadata fields — title tag, meta description, tags, category, slug, and Open Graph text — for a CMS entry, optimized for search discoverability and social sharing.
show-notes-generator
Generates complete, SEO-friendly podcast show notes from a transcript or episode summary — including episode description, key takeaways, timestamps, guest bio, and links section — ready to paste into a podcast host platform.
headline-ab-variants-writer
Takes one original headline and produces five distinct tested variants, each using a different proven approach to drive clicks, shares, or engagement — ready to load directly into a CMS or A/B testing tool.
content-amplifier
Use when the user asks to "amplify influencer content with paid media", "set up whitelisting or Spark Ads", "decide which posts to boost", "repurpose influencer content", "turn one video into multiple ads", or "build a UGC asset library"; produces (paid mode) a content-selection scorecard, a paid amplification…
product-feed-optimizer
Use when the user asks to "optimize my Shopping feed", "fix product disapprovals", "improve product titles/attributes", or "build feed-driven PMax asset groups"; audits and rewrites the Shopping/Performance Max product feed — title/description patterns, required and recommended attributes, GTIN/availability/price…
cold-outbound-sequencer
Use when the user asks to "build a B2B cold-outbound sequence", "design reply-triage branching", "plan a domain warmup / sending throttle", or "make my outbound CAN-SPAM / opt-in compliant"; produces a multi-step outbound sequence with reply-triage branches (positive / objection / referral / not-now / opt-out), a…