Figure Composer

SkillDatabases & data

Give your AI a claim, dataset, or draft result and get back a publication-grade multi-panel scientific figure. It plans the panels, keeps the layout consistent across the figure, and assembles the final version. It can also review the figure along the way so the end result holds up.

Available today. Use it from your connected AI after setup.

Add the skill, then share the claim, dataset, or draft result your figure should present. Your AI will plan the panels and carry the figure through to final assembly.

Then ask your AI: use the Figure Composer skill

What your AI can do with it

  • Plan the panels of a multi-panel scientific figure from a claim or dataset
  • Keep the layout consistent across all panels
  • Review a figure before it is finalized
  • Assemble the final publication-grade figure
  • Work from a claim, a dataset, or a draft result as the starting point

What this skill tells your AI

The instructions your AI receives, as published by companion-inc/feynman in skills/figure-composer/SKILL.md and read by ahel’s review.

Use this skill for multi-panel scientific figures.

Workflow:

  1. State the figure's single claim and list every data artifact that supports it.
  2. Draft a panel plan with labels, chart types, data sources, required annotations, and what each panel proves.
  3. Use figure-style for each plot before composing the multi-panel layout.
  4. Render, inspect, and revise for label collisions, axis readability, color meaning, caption fit, and data-source fidelity.
  5. Save the final figure, source data, generation code, caption, and provenance sidecar.

Do not invent data or hide excluded data in summaries.

Signals

GitHub stars
9k
Forks
1k
Last commit
Sep 2026
Advanced
Catalog kind
skill
Gateway key
figure-composer
Source
github.com/companion-inc/feynman