agentleFS
Yang1BaiYang1Bai/nature-paper-hub/skills/nature-figure/SKILL.md

nature-figure

Skill20 starsChanged 48 days agoMITView on GitHub

Generate publication-quality figures for top-tier journals (Nature, Science, Cell, ACS, Wiley families) using Python (matplotlib) or R (ggplot2). Trigger when user wants to create, polish, or redesign scientific figures for high-impact journals. Handles multi-panel layouts, Nature color palettes, correct typography, and exports SVG/PDF/PNG.

15 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use nature-experiment-log, nature-downloader and 8 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • Delagong950Skill

    nature-experiment-log

    Standardized lab notebook with structured YAML-frontmatter experiment logs. Supports Obsidian vault integration, image/voice/text input, and batch/tracking systems.

    Also in 5 projects that use this

  • vcis-wangchenxuSkill

    nature-downloader

    Help legally obtain, organize, and archive academic papers, PDFs, BibTeX, supplementary files, code links, and benchmark documents for RL, MARL, LLM, and LLM Agent research. Do not bypass paywalls or access controls.

    Also in 5 projects that use this

  • dragonadamSkill

    nature-reviewer

    Simulate a Nature-style reviewer assessment from the referee perspective rather than an author rebuttal. Use when the user wants a pre-submission review, reviewer report, peer-review style critique, novelty/significance/technical soundness assessment, reviewer-style manuscript evaluation, 审稿人视角评估, 预审稿意见, or Nature reviewer report. Return 3 reviewer reports plus a cross-review synthesis, grounded only in the local Nature reviewer source basis.

    Also in 8 projects that use this

  • yanjumlinnb-boopSkill

    nature-paper-to-patent

    Convert scientific papers, theses, technical reports, source code, figures, or research manuscripts into evidence-grounded Chinese invention patent drafts. Use when an AI agent must extract patentable technical contributions, map every claimed feature to source evidence, preserve core formulas as editable Office Math, generate claim-aligned flowcharts and methodology figures, compare a paper with an existing patent, audit support and consistency, or deliver separate Chinese DOCX files for claims, specification, abstract, and abstract figure.

    Also in 5 projects that use this

  • Delagong950Skill

    nature-ref-verifier

    Multi-source cross-validation of academic references. Verify authors, title, year, volume, issue, and pages against PubMed, CrossRef, and publisher pages. Detect fabricated citations.

    Also in 4 projects that use this

  • Yuan1z0825Skill

    nature-shared

    Internal shared-reference support package for installed Nature Skills, including nature-writing, nature-polishing, nature-response, nature-reader, and nature-paper2ppt. Do not invoke it as a standalone user workflow. Load only the specific core or journal-format file requested by another Nature skill.

    Also in 4 projects that use this

The file

We split this skill into the rules it gives an agent

11 rules. 0 are common, 7 are distinctive, and 4 are very specific to this project.

  • 0

    Common

    Most of the 15 edited versions we read keep these.

  • 7

    Distinctive

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    All text must be editable (not rasterized)

    line 31
  2. Distinctive:

    Never use pure red + green together (colorblind conflict)

    line 58
  3. Distinctive:

    Use filled symbols + different shapes for accessibility, not colour alone

    line 59
  4. Distinctive:

    No chartjunk: remove top and right spines

    line 69
  5. Distinctive:

    If user provides a CSV, Excel, or JSON data file:

    line 82
  6. Distinctive:

    Use solid lines with markers.

    line 215
  7. Very specific:

    Prefer horizontal bars for many categories.

    line 218
  8. Very specific:

    Use plt.bar() with edgecolor='black', linewidth=0.5.

    line 219
  9. Very specific:

    Use diverging colourmap (e.g., RdBu_r) for correlation.

    line 226
  10. Distinctive:

    Always include colourbar with label and units.

    line 227
  11. Very specific:

    Always overlay individual data points (stripplot or geom_jitter).

    line 230

“Common” is measured: kept by at least 6 in 10 of the 15 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

nature-figure

Purpose

Generate multi-panel scientific figures that meet Nature portfolio visual standards: correct typography, semantic colour palette, accessible design, and editable SVG output.


Trigger Conditions

Activate when user mentions:

  • "画图" / "figure" / "plot" / "科研绘图"
  • "Nature figure" / "publication figure" / "publication plot"
  • "matplotlib" / "ggplot" / "seaborn"
  • "配色" / "color palette" / "color scheme"
  • Wants to improve or reformat an existing figure

Nature Figure Standards

Typography

  • Font family: Arial or Helvetica (sans-serif, never Times New Roman in figures)
  • Minimum font size in final print: 7 pt (axis labels, tick labels)
  • Panel labels (a, b, c...): 8 pt bold, lowercase
  • Figure title (if any): not embedded in figure — goes in legend
  • All text must be editable (not rasterized)

Size & Resolution

FormatWidthResolution
Single column89 mm (3.5 in)300 DPI min
1.5 column120 mm (4.7 in)300 DPI min
Double column183 mm (7.2 in)300 DPI min
Line artany600 DPI
Final submissionPDF or TIFFvector preferred

Colour Palette (Nature-approved, colorblind-safe)

NATURE_COLORS = {
    "blue":    "#4878CF",
    "red":     "#D65F5F", 
    "green":   "#6ACC65",
    "orange":  "#EE854A",
    "purple":  "#956CB4",
    "teal":    "#82C6E2",
    "brown":   "#D5BB67",
    "gray":    "#8C8C8C",
    # Colorblind-safe primary pair:
    "cb_blue": "#0072B2",
    "cb_orange":"#E69F00",
}
  • Never use pure red + green together (colorblind conflict)
  • Use filled symbols + different shapes for accessibility, not colour alone
  • Grayscale must remain distinguishable

Panel Architecture

  • Each panel makes one clear point
  • Panel (a): overview / schematic / representative image
  • Panels (b–d): quantitative evidence
  • Final panel: comparison or generalizability
  • Panels are labelled a, b, c (lowercase bold, top-left corner)
  • White background; minimal gridlines (light gray, 0.5pt)
  • No chartjunk: remove top and right spines

Statistical Annotations

  • Error bars: always define in legend (mean ± s.d. or ± s.e.m.)
  • Significance: *, **, ***, **** for p < 0.05, 0.01, 0.001, 0.0001; prefer exact p-values
  • n must be stated (e.g., n = 5 independent experiments)
  • Box plots: show median, IQR, whiskers to 1.5×IQR, individual points overlaid

Workflow

Step 0: Auto-figure from data file (fastest path)

If user provides a CSV, Excel, or JSON data file:

python3 ${CLAUDE_PLUGIN_ROOT}/scripts/auto_figure.py \
  --input <data_file> \
  --output ~/Downloads/figure_<date>.pdf \
  --title "[figure title]" \
  --xlabel "X axis label" \
  --ylabel "Y axis label" \
  --type [auto|line|bar|scatter|heatmap|box]

The script auto-detects column types, chooses appropriate chart type, applies Nature style, and saves PDF + PNG.

Step 1: Gather requirements (if no data file yet)

Ask the user:

  1. What data do you have? (paste CSV, describe columns, or share values)
  2. How many panels? What does each panel show?
  3. Single/1.5/double column width?
  4. Python (matplotlib/seaborn) or R (ggplot2)?
  5. Any specific colour requirements or journal sub-style?

Step 2: Generate figure code

Python template (matplotlib):

import matplotlib.pyplot as plt
import matplotlib.gridspec as gridspec
import numpy as np

# --- Nature style settings ---
plt.rcParams.update({
    'font.family': 'Arial',
    'font.size': 8,
    'axes.linewidth': 0.8,
    'axes.spines.top': False,
    'axes.spines.right': False,
    'xtick.major.width': 0.8,
    'ytick.major.width': 0.8,
    'xtick.major.size': 3,
    'ytick.major.size': 3,
    'xtick.direction': 'out',
    'ytick.direction': 'out',
    'figure.dpi': 300,
    'savefig.dpi': 300,
    'savefig.bbox': 'tight',
    'savefig.pad_inches': 0.05,
    'pdf.fonttype': 42,   # editable text in PDF
    'ps.fonttype': 42,
})

COLORS = {
    "blue": "#4878CF", "red": "#D65F5F", "green": "#6ACC65",
    "orange": "#EE854A", "purple": "#956CB4", "teal": "#82C6E2",
    "cb_blue": "#0072B2", "cb_orange": "#E69F00",
}

# --- Figure layout ---
fig = plt.figure(figsize=(7.2, 4.0))  # double column, adjust height
gs = gridspec.GridSpec(1, 3, figure=fig, wspace=0.4, hspace=0.4)

ax_a = fig.add_subplot(gs[0])
ax_b = fig.add_subplot(gs[1])
ax_c = fig.add_subplot(gs[2])

# --- Panel labels ---
for ax, label in zip([ax_a, ax_b, ax_c], ['a', 'b', 'c']):
    ax.text(-0.15, 1.05, label, transform=ax.transAxes,
            fontsize=8, fontweight='bold', va='top', ha='right')

# --- YOUR DATA GOES HERE ---
# ax_a: ...
# ax_b: ...
# ax_c: ...

plt.savefig('figure1.pdf', format='pdf')
plt.savefig('figure1.png', dpi=300)
print("Saved: figure1.pdf, figure1.png")

R template (ggplot2):

library(ggplot2)
library(patchwork)

# Nature theme
theme_nature <- function() {
  theme_classic(base_size = 8, base_family = "Arial") +
  theme(
    axis.line = element_line(linewidth = 0.5),
    axis.ticks = element_line(linewidth = 0.5),
    axis.ticks.length = unit(2, "pt"),
    strip.background = element_blank(),
    legend.key.size = unit(3, "mm"),
    plot.margin = margin(2, 2, 2, 2, "mm")
  )
}

nature_colors <- c(
  blue = "#4878CF", red = "#D65F5F", green = "#6ACC65",
  orange = "#EE854A", purple = "#956CB4", teal = "#82C6E2"
)

# --- YOUR PLOTS ---
# p1 <- ggplot(...) + theme_nature()
# p2 <- ggplot(...) + theme_nature()
# combined <- p1 | p2
# ggsave("figure1.pdf", combined, width = 183, height = 80, units = "mm", dpi = 300)

Step 3: Validate

Before outputting, check:

  • Font ≥ 7pt in all elements
  • No top/right spines
  • Colorblind-safe palette used
  • Error bars defined
  • Panel labels present (a, b, c lowercase bold)
  • Resolution ≥ 300 DPI (600 for line art)
  • Figure width matches column format
  • PDF/SVG output for vector editability

Step 4: Figure legend

Generate the corresponding figure legend text:

  • Bold "Figure X |" prefix
  • Short title (one phrase)
  • One sentence per panel
  • Error bar definition
  • n values and statistical test used
  • Scale bar definition (for images)

Common Figure Types

Line plot (time series / trends)

Use solid lines with markers; different line styles + colours for groups.

Bar chart

Prefer horizontal bars for many categories; overlay individual data points. Use plt.bar() with edgecolor='black', linewidth=0.5.

Scatter plot

Include regression line with 95% CI if showing correlation. State Pearson/Spearman r and p-value on plot.

Heatmap

Use diverging colourmap (e.g., RdBu_r) for correlation; sequential for intensity. Always include colourbar with label and units.

Box/Violin plot

Always overlay individual data points (stripplot or geom_jitter). State n per group.

Schematic / Mechanism diagram

Recommend using BioRender (biorender.com) or Inkscape for schematics. Export as SVG and embed in figure.


Output

Provide the user with:

  1. Complete, runnable Python or R code
  2. Instructions to save in the correct format (PDF + PNG)
  3. The figure legend text
  4. A checklist of what to verify before submission

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.