Is there a Claude Code skill that makes infographics?
Yes, Wireflow's infographic skill for Claude Code makes two infographics from a brief, stats or an article, with text spelled exactly, for 38 credits a run.
Yes, Wireflow's ai-infographic-generator skill turns a brief, a set of stats or an article into an infographic inside Claude Code. Claude picks the structure, writes short copy for every section and lets you edit each word, then quotes the credits and waits for your yes. The skill's Wireflow app then draws the same text in two portrait layouts at 1728x2304 with Ideogram 4.5, and Claude saves both PNGs to ./wireflow-outputs/.
The thinking happens in Claude, and the drawing runs as a published Wireflow app over the Wireflow MCP connector, signed in as you. Without the connector, the skill still writes the copy deck and a layout sheet a designer can build. You can read the skill itself in its folder on GitHub.
Use it when you need an infographic, an explainer graphic, a process or step diagram, a comparison chart, a one-page visual summary of a blog post, or a stats graphic for a client report or social post. For feature graphics on a product listing, made from a product photo, use the Amazon product images skill. For ad concepts and headlines rendered as static ads in 1:1, 4:5 and 9:16, use the ad creative skill.
How do I install the infographic skill in Claude Code?
Install the Wireflow plugin, which adds every Wireflow skill plus the Wireflow connector:
/plugin marketplace add wireflowINC/skills
/plugin install wireflow-skills@wireflow
Claude Code first shows the plugin's details, so you can review it and choose its scope. Then run /mcp, choose wireflow and sign in with your Wireflow account. The skill shows up as /wireflow-skills:ai-infographic-generator, and Claude also picks it up from plain requests.
To add only this skill, use the skills CLI, or copy the skills/ai-infographic-generator folder into ~/.claude/skills/ (all projects) or .claude/skills/ (one project). Either way, connect Wireflow yourself, then run /mcp to sign in. The connector uses OAuth, so there is no API key to copy.
npx skills add wireflowINC/skills --skill ai-infographic-generator
claude mcp add --transport http wireflow https://www.wireflow.ai/api/mcp
To check the setup, ask Claude for an infographic. The skill's first step is a connection check for the Wireflow tools run_app, get_execution, get_credit_balance and whoami. If they are missing, it says so in one sentence, gives you the setup steps and still writes the copy deck and layout sheet.
What do I give Claude to make an infographic?
Give Claude your material as text: a brief, a pasted article or its key points, or the figures along with where each one comes from. The skill also asks who will see the graphic and where (a client report, LinkedIn, a blog header or a slide), plus your brand colors, any footer or logo text, and anything that must appear word for word.
A request might look like this:
Make an infographic for our agency's LinkedIn: how we turn one blog post
into a week of social content. Title: "1 Blog Post, 7 Days of Content".
Steps: pull 5 key points, write one post per point, design a quote card,
cut a short clip, schedule the week. Brand colors #12355B and #F2A900,
footer "example.com".
Every number on the graphic has to come from you or from the source you gave, and if a figure has no source, Claude asks about it or leaves it out. A logo file can't go in, because the app is text only: a brand name can only appear as typed text. The skill also won't invent data or sources, or present third-party logos or trademarks as endorsements.
Claude then plans the graphic with you:
- It picks the structure that fits your material and says why in one line: process (ordered steps), comparison (two options side by side), list (tips or reasons), timeline (dates) or data (3 to 6 figures with labels).
- It writes the copy deck: a title of 8 words or fewer and 3 to 6 sections, each with a label of 2 to 5 words and one short line of 8 words or fewer, plus an optional source line and footer.
- It proposes two layouts that suit the structure and differ clearly, such as a numbered vertical list and a card grid for a process, or a bar chart and big-number tiles for data.
- It shows the deck and both layouts as a short table, and you can change any word.
- It quotes the credits, and nothing renders until you say yes to that deck and that price.
On every run, Claude fills in all five inputs of the skill's Wireflow app:
| Input | What goes in it |
|---|---|
topic |
What the graphic is about, as context. It is not drawn unless it is also in infographic_text |
infographic_text |
Every word on the graphic, each piece in double quotes, in reading order: title, section labels with their short lines, figures, source, footer |
style |
Palette (hex codes if known), font feel and icon style, shared by both layouts |
layout_1, layout_2 |
How the sections are arranged, one arrangement each, such as vertical list, card grid, timeline, two columns, bar chart or cycle |
What do I get back from a run?
You get two PNG infographics at 1728x2304 (3:4 portrait), with the same text drawn in two different layouts. Both have a flat vector look in one shared style and show only your quoted words, with no invented numbers, labels, logos or photos.
Claude starts one run per approved deck with run_app, then polls get_execution every 5 seconds while the status is PENDING, RUNNING or IN_PROGRESS. In the skill's own test, a run took under a minute. If a run ends as FAILED, CANCELLED or TIMEOUT, Claude stops and quotes the error.
Claude saves each file under ./wireflow-outputs/YYYY-MM-DD/infographics/, named after the topic and the layout, and reads every image before handing it over. It checks that:
- each word and number matches the approved deck letter by letter, with nothing missing and nothing extra
- the sections are in the right order and each line sits under its own label
- the icons fit their sections (a wrong or missing icon is common, so Claude names it)
- arrows, connectors and chart bars match the layout you asked for and don't imply numbers that aren't in the text
It lists every miss per file, then closes with a table of each file, its layout and its CDN link, plus the credits spent (your balance before minus after). It also offers a second version with edited copy or two other layouts, as one more run with its own quote, and the skill's board in Wireflow, where you can change the size or add steps. This app does not render square or landscape crops; for those, the skill suggests giving the layout sheet to a designer.
Will it spell my text correctly?
It is built to spell every word as given: the app draws only the quoted words from your approved deck, with Ideogram 4.5, a model that keeps lettering clean, and MagicPrompt off so nothing rewrites your copy.
Fewer words render more reliably, and the deck keeps the title and each section line to 8 words or fewer. Settle the wording before the render: you can change any word in the deck table, but after rendering, a text change means a new run.
Claude still compares every word and number in each image with the approved deck, letter by letter, and lists every miss per file. A fix is a new run of 38 credits, and it starts only after your yes.
How many credits does one run cost?
One run costs 38 credits and makes both layouts, which is 19 credits per image, as measured with Ideogram 4.5 on 19 October 2026.
| What | Credits |
|---|---|
| Structure, copy deck and the quote | 0 |
| One run: two PNGs at 1728x2304 | 38 |
| Each image in that run | 19 |
| A new version or a fix | 38 per run, only after your yes |
Before rendering, Claude calls whoami and get_credit_balance, then quotes the run in a line like "1 run x 38 credits = 38 credits for 2 layouts" and shows your balance. If your balance is short, it stops and says so.
Rendering needs a Wireflow account with credits. See pricing for what each plan includes.
Does the infographic skill work in Cursor, Codex or Claude Desktop?
It is built for Claude Code, the skills CLI also installs it in Cursor and Codex, and its own setup steps cover Claude.ai and Claude Desktop through a custom connector. Whichever you use, that agent must be connected to the Wireflow MCP server before the skill can render anything.
In Cursor, Codex, Gemini CLI and other agents that read SKILL.md, install it with the npx skills add command above. Wireflow says it has not tested its skills in other agents itself. For connection steps, see the guides for Cursor and Codex, or Use Wireflow from your agent for every client.
In Claude.ai or Claude Desktop, add https://www.wireflow.ai/api/mcp as a custom connector. Saving files needs a shell. Without a Wireflow connection, the skill still writes the copy deck and the layout sheet for a designer to build.
More Wireflow skills, and how they all work, are on the agent skills page.
Related
- Ad creatives
Yes, Wireflow's ad creatives skill for Claude Code writes concepts and headlines, then renders each one as 1:1, 4:5 and 9:16 images in one call for 60 credits.