agentleFS
LocloveLoclove/cumcm-skills/skills/cumcm-figure/SKILL.md

cumcm-figure

Skill4 starsChanged 3 months agoMITView on GitHub

Create, revise, and audit polished scientific figures that visually merge with Song-font Chinese CUMCM papers, including model comparison, time series, spatial plots, sensitivity analysis, residual diagnostics, feature importance, flowcharts, claim-driven captions, body-text anchors, and source-data linkage. Use when the user asks for beautiful plots,科研绘图,论文图表,可视化,figure export, or figure-review for a modeling paper.

The file

We split this skill into the rules it gives an agent

12 rules. About 0 are common, 7 are distinctive, and 5 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈7

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 5

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    If the user asks for figures4papers style, Nature-style AI-paper visuals, multi-panel comparison bars, radar/trend panels, heatmaps, or a …

    line 16
  2. Very specific:

    Use scripts/figures4papers_adapter.py index --repo <figures4papers_path> to find matching upstream demo patterns when a local clone exists.

    line 17
  3. Very specific:

    Use scripts/figures4papers_adapter.py scaffold --kind <kind> --out <script.py> to generate a publication-style plotting scaffold, then …

    line 18
  4. Very specific:

    Use scripts/plot_template.py for simple CSV-based charts or write a task-specific script when the figure requires custom logic.

    line 19
  5. Distinctive:

    Check axis labels, units, legends, color contrast, formula symbols, captions, source-data linkage, and all typography at final paper size.

    line 21
  6. Very specific:

    Keep figure Chinese text in Song-style fonts, English/numbers in Times New Roman where available, and terminology/precision/color identity …

    line 22
  7. Distinctive:

    Do not make a figure before the claim and data source are known.

    line 47
  8. Distinctive:

    Do not hide uncertainty, missing data, or poor model fit.

    line 48
  9. Distinctive:

    Do not use unexplained acronyms, unlabeled axes, or image-only formulas in captions.

    line 49
  10. Distinctive:

    Do not use a figure as decoration or leave it unmentioned/uninterpreted in the body.

    line 50
  11. Distinctive:

    Do not mix Song, YaHei, SimHei, Kaiti, and FangSong casually across the paper and figures.

    line 51
  12. Distinctive:

    Do not copy figures4papers source scripts, figures, or paper artwork wholesale into a contest project.

    line 52

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “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

CUMCM Figure

This skill creates figures that support claims, not decorative charts.

It can also use figures4papers as an external template/style base for polished matplotlib figures. The adapter does not vendor the upstream repository's source code or image assets; it indexes and references local or cloned figure_* demos, then scaffolds CUMCM-specific plotting scripts with compatible publication-style conventions.

Workflow

  1. Define the claim each figure must prove and the body paragraph that will introduce it.
  2. Load references/figure-standards.md and ../cumcm-shared/references/enhanced-writing-profile.md for style, typography, paper placement, and export rules.
  3. If the user asks for figures4papers style, Nature-style AI-paper visuals, multi-panel comparison bars, radar/trend panels, heatmaps, or a polished matplotlib script, load references/figures4papers-adapter.md.
  4. Use scripts/figures4papers_adapter.py index --repo <figures4papers_path> to find matching upstream demo patterns when a local clone exists.
  5. Use scripts/figures4papers_adapter.py scaffold --kind <kind> --out <script.py> to generate a publication-style plotting scaffold, then adapt it to the user's data.
  6. Use scripts/plot_template.py for simple CSV-based charts or write a task-specific script when the figure requires custom logic.
  7. Export at least one vector format (.pdf or .svg) for paper assembly and one raster preview (.png) when helpful.
  8. Check axis labels, units, legends, color contrast, formula symbols, captions, source-data linkage, and all typography at final paper size.
  9. Keep figure Chinese text in Song-style fonts, English/numbers in Times New Roman where available, and terminology/precision/color identity identical to the paper.
  10. Return a body anchor before the figure and an interpretation paragraph after it; hand off captions to cumcm-paper-writing and traceability to cumcm-data-loop.

Output Contract

Figure plan
| Figure | Claim | Data source | Chart type | Export | Caption point |

Generated files
- ...

Figure QA
- Units:
- Color/legend:
- Readability:
- Source data:
- Paper placement:
- Body anchor before figure:
- Interpretation after figure:
- Typography match:

Red Lines

  • Do not make a figure before the claim and data source are known.
  • Do not hide uncertainty, missing data, or poor model fit.
  • Do not use unexplained acronyms, unlabeled axes, or image-only formulas in captions.
  • Do not use a figure as decoration or leave it unmentioned/uninterpreted in the body.
  • Do not mix Song, YaHei, SimHei, Kaiti, and FangSong casually across the paper and figures.
  • Do not copy figures4papers source scripts, figures, or paper artwork wholesale into a contest project. Use them as style references and cite/attribute when a figure design materially follows an upstream paper example.

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.