Publication Figures logo

Publication Figures

OrganizationPopular
ai4s-research
publication-figures

Use whenever you generate or review a chart, plot, table, or paper figure in this workspace, including work delegated by paper-writing, literature-survey, and experiment skills. Applies the Open Science publication style, enforces readable final-size layout for figures and tables, and rejects generic diagram-tool output as a publication figure. Interactive Plotly/HTML may be used for exploration, but paper delivery requires a static publication-ready export.

Overview

Publisherai4s-research
Repositoryopen-science
Skill namepublication-figures
Stars
1.7K
Forks
201
Bundled files
1
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 1 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by ai4s-research on GitHub. Read the source before you install it.

Installation

Install the Publication Figures AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/ai4s-research/open-science.git /tmp/open-science
mkdir -p .claude/skills
cp -r /tmp/open-science/runtime/skills/core/publication-figures .claude/skills/publication-figures
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Publication Figures in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Publication Figures on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Publication Figures is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Publication Figures and Tables

Make generated figures publication-grade and on-system by default. Every figure you produce with matplotlib must use the bundled Open Science style, so a figure in a report and a stat tile in the app read as one design system. When a paper/survey skill or target venue specifies its own publication palette and physical type scale, that more specific standard overrides this palette; the final-size, collision, and figure-form rules below still apply.

Apply the style (always, before plotting)

The style file openscience.mplstyle sits next to this SKILL.md. Load it by absolute path at the top of any figure script:

python
import matplotlib.pyplot as plt
from pathlib import Path

# This skill's directory — the style ships beside SKILL.md.
STYLE = Path(__file__).resolve().parent / "openscience.mplstyle" if "__file__" in dir() else None
# In a notebook/agent cell, use the skill's deployed path directly:
plt.style.use(str(STYLE)) if STYLE and STYLE.exists() else plt.style.use("default")

If you cannot resolve the path, set the palette inline (same hexes as below).

Choose a paper-appropriate form

  • Use matplotlib/seaborn for quantitative evidence; use a purpose-built TikZ/SVG/PDF schematic for a method or architecture; use scientific image panels only with source, crop, scale, and processing provenance.
  • Do not use Mermaid, PlantUML, generic flowchart/mind-map output, diagram-editor screenshots, or notebook/UI screenshots as a final paper figure. They may be scratch aids only. Graphviz is acceptable only when the graph itself is the analysed data, not as a shortcut for a generic process diagram.
  • Every figure must make evidence, a mechanism, or an experimental design easier to understand. Decorative roadmaps, funnels, icon collages, and stock process diagrams do not belong in a paper.

The shared palette (single source of truth)

These are the exact hues the app's native charts use. Assign categorical series in this fixed order — never a different order, never a cycled 9th hue.

SlotHueLight hex
1blue#2a78d6
2aqua#1baf7a
3yellow#eda100
4green#008300
5violet#4a3aa7
6red#e34948
7magenta#e87ba4
8orange#eb6834

Sequential (magnitude, one hue light→dark): #cde2fb #9ec5f4 #6da7ec #3987e5 #256abf #184f95 #104281. Diverging: blue ↔ red with a neutral gray midpoint.

Rules (from the app's dataviz standard)

  • One y-axis. Never two scales on one plot — use two charts or index to a common base.
  • Categorical color = identity, assigned in slot order; sequential = one hue by magnitude; diverging = two hues + gray midpoint. Never a rainbow.
  • Thin marks, recessive chrome: 2px lines, ≥6pt markers, hairline y-grid only, no top/right spines (the style sets these).
  • Label selectively — the endpoint or the extreme, never a number on every point. A legend is present for ≥2 series; a single series needs none (the title names it).
  • Text stays in ink, never the series color. Identity comes from the mark.
  • Fit by design, not tiny type. Size for the final column/page first. When labels compete for space, enlarge the canvas, wrap/shorten labels, show fewer ticks, move the legend, use a horizontal chart, or split panels. Do not shrink any text below 7 pt at final size.
  • No collisions. Labels, ticks, legends, annotations, watermarks, panel letters, and data marks may not overlap or be clipped. Use constrained layout where appropriate, but always inspect the rendered result; an automatic layout call is not proof.
  • Save clean: plt.savefig(path, bbox_inches="tight") (the style sets dpi).

Keep paper tables readable

  • Design every table for its final column or page width. Body text must remain at least 8 pt at final size.
  • Resolve an over-wide table in this order: shorten or wrap headings and move units/details into the caption or notes; use flexible text columns and aligned numeric columns; split the table or move secondary columns to an appendix; use landscape for a genuinely wide appendix table; reduce cell padding last.
  • Do not solve width by scaling the entire table until it fits. If scaling would make text smaller than 8 pt, redesign or split the table.
  • Keep precision purposeful, repeat units once, and avoid dense vertical rules.
  • Render the containing PDF or page to an image at the intended output size and inspect it. A table fails if it crosses margins, clips content, overlaps text, or is readable only by zooming.

Final-size inspection

For PDF output, rasterize the actual final artifact before delivery:

bash
pdftoppm -png -r 180 paper.pdf /tmp/paper-check/page

Inspect every page containing a generated figure or table. Source dimensions, tight_layout, constrained_layout, and a successful compile are not evidence that the rendered result is readable.

Checklist before returning a figure or table

  1. Style applied (palette + chrome from openscience.mplstyle).
  2. Series colors assigned in slot order; ≤8 series (else group into "Other").
  3. Single y-axis; legend iff ≥2 series; axis labels + units present.
  4. Figure form is appropriate for a paper; no Mermaid/PlantUML/generic diagram.
  5. Table fits by structure rather than tiny type; final body text is at least 8 pt.
  6. Final PDF/page was rasterized and visually inspected: readable text, no overlap, clipping, or margin overflow.
  7. Saved to the workspace and referenced by path so it surfaces as an artifact.

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Publication Figures AI skill do?

Use whenever you generate or review a chart, plot, table, or paper figure in this workspace, including work delegated by paper-writing, literature-survey, and experiment skills. Applies the Open Science publication style, enforces readable final-size layout for figures and tables, and rejects generic diagram-tool output as a publication figure. Interactive Plotly/HTML may be used for exploration, but paper delivery requires a static publication-ready export.

Why use Publication Figures on TypingMind?

Because you install it once and use it with any model. Publication Figures is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Publication Figures in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/ai4s-research/open-science/tree/master/runtime/skills/core/publication-figures. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Publication Figures?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Publication Figures?

As many as you like. As long as a model supports skills, you can use Publication Figures with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Publication Figures AI skill free?

It is published on GitHub by ai4s-research. Check the repository for licensing terms. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇