og-image-design — Claude Code skill: OG cards, what it does, install
og-image-design builds the image behind a link preview: per-network sizes, text layout rules, and rendering through the inference.sh CLI.
no license 244 installs ★ 31 updated 2026-08-31
npx skills add 101-skills/superpowers --skill og-image-design -a claude-codeWhat og-image-design does
The skill collects what a link preview needs. You get image sizes for Facebook, Twitter/X and LinkedIn, plus limits on headline length and padding from the edges. A fallback that works everywhere: 1200 x 630 px, PNG or JPG, under 5 MB. Dark cards stand out in feeds that are mostly white, and the templates here are dark.
Nothing gets drawn by hand. A block of HTML goes into the html field and the infsh/html-to-image app renders it through the belt CLI. For a cover with no text, falai/flux-dev-lora takes a prompt and the 1200 x 630 dimensions. Headline, subtitle and gradient all sit in the markup.
belt login comes first; the run commands assume an authenticated CLI. Tool access is pinned to Bash(belt *), so the skill stays inside belt. There is also a meta tag reference: og:title, og:description, og:image, og:url, and Twitter's summary_large_image card.
What it helps with
- turn one layout into previews for Facebook and LinkedIn
- check whether a headline fits the character limit
- render a 1200 x 630 card without a design tool
- wire up
og:imageandtwitter:cardon a page - pick text colors that pass the WCAG AA contrast bar
Example prompts for the agent
- Make an OG image for my post on build times: dark gradient, headline under 60 characters, 1200 x 630.
- Audit the
og:titleandtwitter:cardtags on this page for LinkedIn previews. - Generate a text-free cover with
falai/flux-dev-loraat 1200 x 630.
How to install og-image-design
1Skills CLI (Vercel)
npx skills add 101-skills/superpowers --skill og-image-design -a claude-code2Manually: put the files in place
~/.claude/skills/og-image-design/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/og-image-design/
1Skills CLI (Vercel)
npx skills add 101-skills/superpowers --skill og-image-design -a codex2Manually: put the files in place
.agents/skills/og-image-design/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add 101-skills/superpowers --skill og-image-design -a cursor2Manually: put the files in place
.cursor/skills/og-image-design/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install 101-skills/superpowers guides/design/og-image-design2Manually: put the files in place
.github/skills/og-image-design/SKILL.md1Skills CLI (Vercel)
npx skills add 101-skills/superpowers --skill og-image-design -a gemini-cli2Platform CLI
gemini skills install https://github.com/101-skills/superpowers --path guides/design/og-image-designManually: put the files in place
.devin/skills/og-image-design/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add 101-skills/superpowers --skill og-image-design -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about og-image-design
How do I install og-image-design in Claude Code?
Run npx skills add 101-skills/superpowers --skill og-image-design -a claude-code. The skill file ends up at ~/.claude/skills/og-image-design/SKILL.md.
What does og-image-design do?
It guides the making of a social sharing image: network-specific sizes and limits, layout rules for text, and ready commands that render the card through infsh/html-to-image.
What do I need before generating an image?
The belt CLI from inference.sh and a belt login session. The CLI itself comes from npx skills add belt-sh/cli, and belt app run will not go through without signing in.