Skip to content
Back to templates

data-faithful-infographic

@randalolson/data-faithful-infographic

Use when you have a data story for LinkedIn, X, or a blog and want a designed square infographic whose visible numbers, labels, and source line exactly match your data, with no invented figures, transposed digits, or decorative chart-junk.

Outcome

You go from a data story idea to a finished square infographic, ready to post on LinkedIn, X, or a blog, with every visible number and category label faithful to the source data you supplied.

See it in action

You hand the workflow a data story in plain language. It pulls out the numbers, confirms them with you, and checks every figure it renders against that confirmed list before the image ever reaches you. The run below uses one such story, the world's most spoken languages, with and without that check.

Without the workflow

A busy infographic crowded with a dragon, national flags, and landmarks, with an altered title

One generation, nothing holding it to the data. The model chases the look of an infographic and loses the substance. It rewrote the title, dropped the source line, drew flags for languages that are not in the dataset, and packed the rest behind a dragon and a row of landmarks. The verifiers flag every one of those:

CheckVerdict
Number fidelityFAIL, the title became "most spoken native languages" and the source was swapped for a vague "approximate" note
IconographyFAIL, it added a French speech bubble and a Romanian flag, languages not in the data
No chart-junkFAIL, a dragon, landmarks, and mascots crowd out the numbers

With the workflow

A clean bar chart with Mandarin Chinese 939M leading, then Spanish 485M, English 380M, Hindi 345M, and Portuguese 236M in distinct colors

Same brief, same model, this time told which numbers it has to honor and rechecked until it does. The figures match, Mandarin clearly leads, and the decoration is gone. The image ships only after all six checks pass:

CheckVerdict
Number fidelity, every figure matches the briefPASS
One dominant claimPASS
Truthful scalingPASS
No chart-junkPASS
Legible, no overlapPASS
Iconography, no invented categoriesPASS

Use this template

Your agent fetches this runbook and runs it, revising the output until the verifiers pass.

First time with Goodeye?

Connect your agent over MCP, then ask it to fetch this template by its identifier. Or install the CLI and run the command below.

Connect your agent (Claude Code)
claude mcp add --transport http goodeye https://mcp.goodeye.dev/mcp
Or install the CLI
uv tool install goodeye

Connecting over MCP prompts a quick sign-in. The CLI can fetch and run a public template with no account.

Set up another clientFull quickstart

Fetch or fork

Fetch with CLI
goodeye templates get @randalolson/data-faithful-infographic
Fork with CLI
goodeye templates fork @randalolson/data-faithful-infographic

Also available from

MCP
get_template(identifier="@randalolson/data-faithful-infographic")
REST
curl https://api.goodeye.dev/v1/templates/@randalolson/data-faithful-infographic