Social Thumbnail Workflow

Social Thumbnail Workflow is a skill for Claude Code from jau123/MeiGen-AI-Design-MCP. It costs 123 tokens per session (1,063 once invoked), scanned A, original, MIT.

A workflow for creating vertical images for short-video platforms and mobile social feeds, where content is usually viewed in a 9:16 portrait layout.

In plain words
What is it for?
Use it for TikTok, Reels, and YouTube Shorts covers, video thumbnails, and other vertical social-media posters.
Why use it?
It keeps the main subject and headline readable while people quickly scroll through phone-sized content.

Skill for Claude Code

Written for Claude Code: shipped in a Claude Code plugin. Also seen: names the AskUserQuestion tool.

Part of the meigen plugin — 3 skills, 4 commands shipped together

Good fit Use it for TikTok, Reels, and YouTube Shorts covers, video thumbnails, and other vertical social-media posters.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/jau123/meigen-ai-design-mcp/social-thumbnail
About the project

MeiGen AI Design MCP is an MCP server that lets AI coding tools generate images and videos through cloud models, OpenAI-compatible APIs, or local ComfyUI, an interface for running image-generation workflows. It is for creating designs such as logos, product images, animations, and batches of visual variations from coding-agent environments. Its catalogue entries provide skills, commands, agents, a plugin, and a hook for using the server's design workflow.

jau123/MeiGen-AI-Design-MCP · 1,749 stars · on GitHub · meigen.ai

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 jau123/MeiGen-AI-Design-MCP --skill social-thumbnail
Clone the repo
git clone --depth 1 https://github.com/jau123/MeiGen-AI-Design-MCP

Made for: Claude Code.

Or install meigen, the plugin that ships this one along with the rest of its 3 skills, 4 commands.

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 Social Thumbnail Workflow

README.md
[![agentmods](https://agentmods.dev/badge/skills/jau123/meigen-ai-design-mcp/social-thumbnail/github.svg)](https://agentmods.dev/skills/jau123/meigen-ai-design-mcp/social-thumbnail)
Your own site
<a href="https://agentmods.dev/skills/jau123/meigen-ai-design-mcp/social-thumbnail"><img src="https://agentmods.dev/badge/skills/jau123/meigen-ai-design-mcp/social-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.

agentmods 80×15 button for Social Thumbnail Workflow

Your own site · 80×15
<a href="https://agentmods.dev/skills/jau123/meigen-ai-design-mcp/social-thumbnail"><img src="https://agentmods.dev/badge/skills/jau123/meigen-ai-design-mcp/social-thumbnail.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 123 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,063 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector pass 7 Sept 2026
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 $0.00123 $0.01063
Opus 5 $0.00062 $0.00531
Sonnet 5 $0.00025 $0.00213
Haiku 4.5 $0.00012 $0.00106

Measured 12d ago against content hash cab3c2d2434f, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

Social Thumbnail Workflow 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.

plugin/skills/social-thumbnail/SKILL.md · 78 lines

How it starts

The opening of the file, as written. The whole thing — 78 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Social Thumbnail Workflow

Generate cover art for short-video platforms (TikTok / Reels / Shorts) or vertical social feeds. Optimized for 9:16, high contrast, big-headline space, and 1-second feed legibility.

When to trigger

  • User says "video thumbnail", "封面", "短视频封面", "Reels cover", "TikTok cover"
  • User mentions "9:16" or "vertical poster" or "竖版海报"
  • User describes content meant for mobile feeds and asks for visual

Design constraints (apply to every prompt)

  1. 9:16 aspect ratio — pass aspectRatio: "9:16" explicitly to every generate_image call. Do NOT write --ar 9:16 in the prompt.
  2. High contrast — bold foreground subject against a strong color-block or gradient background. Tiny details get lost when scaled to a phone-feed thumbnail.
  3. Headline-friendly composition — leave a clear zone (top third or bottom third) for the user's title text. Mention this explicitly in the prompt: "leave top third clear for headline overlay".
  4. One focal point — feed thumbnails work when the eye lands in <1 second. Don't pack multiple subjects.
  5. Brand-safe color — when the user gives a brand color, pin it; otherwise default to high-saturation primary palettes (red / cobalt / orange / electric green) over muted tones.

Workflow

1. Clarify intent

Ask ONE question (per UX Rule 4 — don't batch-ask):

  • "What's the headline text the cover needs to support?" (lets you reserve the right zone)
  • OR if user already gave a clear theme: skip ask, plan directly.

2. Reference (optional but recommended)

Call search_gallery(category="Poster Design") for style reference; 146 curated entries are available. Show 3-5 thumbnails to the user if they want to pick a direction.

3. Plan 2-3 variants (default 3)

Distinct directions, not minor tweaks. For each, write a prompt that:

  • Names the focal subject (1 person / 1 object / 1 typographic motif)
  • Specifies background (color block / gradient / minimal scene)
  • Specifies lighting / mood
  • Calls out the headline-safe zone explicitly
  • Includes contrast keyword ("high contrast", "bold colors")

Read the full file on GitHub · 78 lines

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. 12d ago First seen · 78 lines · 123 tokens per session scan A cab3c2d2434f

Subscribe to this mod's changes

Social Thumbnail Workflow is a skill published in the GitHub repository jau123/MeiGen-AI-Design-MCP (1,749 stars, last pushed 1mo ago), licensed MIT. It adds 123 tokens to every session and 1,063 once invoked, about $0.0006 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 2026-08-30.