agentleFS
YulivuYulivu/debuffer-skills/skills/paper-visualization/SKILL.md

paper-visualization

Skill9 starsChanged 3 months agoMITView on GitHub

Route paper figure work to the right lightweight path and enforce manuscript-grade visual rules. Use when user says \"论文作图\", \"补图\", \"论文可视化\", \"画论文图\", \"paper visualization\", \"figure pipeline\", or wants one entry point for plots, architecture figures, and figure audits.

The file

We split this skill into the rules it gives an agent

9 rules. About 0 are common, 4 are distinctive, and 5 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    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. Distinctive:

    Vector first: prefer PDF / SVG / native LaTeX vector output over raster.

    line 34
  2. Very specific:

    Local-lightweight default: do not invent heavyweight GUI workflows when paper-figure, figure-spec, or template-backed LaTeX/matplotlib can …

    line 38
  3. Very specific:

    Use /paper-illustration or /paper-illustration-image2 only when the figure is not naturally expressed as data plots or deterministic …

    line 66
  4. Very specific:

    Use /mermaid-diagram only for low-stakes drafts, planning, or quick visual communication.

    line 70
  5. Very specific:

    After generation, run /figure-table-audit on the final figure set.

    line 74
  6. Distinctive:

    When a user asks for "paper figures" without enough detail:

    line 78
  7. Very specific:

    Read PAPER_PLAN.md, paper/, figures/, and any existing figure plan.

    line 80
  8. Distinctive:

    Do not stop at "use skill X".

    line 118
  9. Distinctive:

    if the user asked only for process guidance, return the routed figure plan and the typography/export constraints.

    line 120

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 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

Paper Visualization

Capability Routing

This is a first-layer entry skill. Keep it loaded as the user-facing route; when a request needs a specialized capability below, resolve the debuffer repo root from .debuffer_skills/installed-skills-codex.txt (repo_root) when available, read the referenced library SKILL.md, then follow that skill. Do not copy the whole library skill into this file.

  • /paper-figure: read ../library/paper/paper-figure/SKILL.md.
  • /figure-spec: read ../library/paper/figure-spec/SKILL.md.
  • /image-to-vector-ppt: read ../library/paper/image-to-vector-ppt/SKILL.md.
  • /figure-table-audit: read ../library/paper/figure-table-audit/SKILL.md.
  • /mermaid-diagram: read ../library/paper/mermaid-diagram/SKILL.md.
  • /render-html: read ../library/paper/render-html/SKILL.md.
  • /paper-illustration: read ../library/paper/paper-illustration/SKILL.md.
  • /paper-figure-artifact-audit: read ../library/paper/paper-figure-artifact-audit/SKILL.md.

Route paper-figure work for: $ARGUMENTS

This is a lightweight orchestration skill. It does not replace the existing renderers. It decides which skill should own each figure and applies one shared publication rule set before any drawing starts.

Shared Rules

  • Body-font lock: any visible text inside a figure must match the paper body font size unless the venue explicitly requires a documented exception.
  • Vector first: prefer PDF / SVG / native LaTeX vector output over raster.
  • No title inside figure: titles belong in captions, not inside the image.
  • Reproducible artifacts: save the generating script/spec next to the output.
  • Local-lightweight default: do not invent heavyweight GUI workflows when paper-figure, figure-spec, or template-backed LaTeX/matplotlib can do the job.

Default Routing

Choose the narrowest route that matches the requested figure:

  1. Data plots and result figures
    Use /paper-figure. Typical cases: line/bar/scatter/heatmap/box plots, ablations, scaling curves, comparison tables, multi-panel result figures.
  2. Architecture / workflow / system / pipeline diagrams
    Use /figure-spec first. Choose this when layout is structured and should stay deterministic, editable, and easy to revise.
  3. Native LaTeX vector figures
    Stay in the project with PGFPlots/TikZ when:
    • the figure should inherit manuscript typography exactly,
    • the figure is best expressed as axis-based LaTeX graphics,
    • or the target venue is sensitive to font consistency. Prefer:
    • templates/figure/matplotlib_publication_style.py
    • templates/figure/pgfplots_bodyfont.tex
  4. Qualitative concept illustrations
    Use /paper-illustration or /paper-illustration-image2 only when the figure is not naturally expressed as data plots or deterministic diagrams.
  5. Simple temporary flowcharts
    Use /mermaid-diagram only for low-stakes drafts, planning, or quick visual communication. Do not treat Mermaid as the default for submission figures.
  6. Submission-quality check
    After generation, run /figure-table-audit on the final figure set.

Triage Procedure

When a user asks for "paper figures" without enough detail:

  1. Read PAPER_PLAN.md, paper/, figures/, and any existing figure plan.
  2. Build a compact figure inventory:
    • figure id
    • purpose / claim supported
    • preferred route
    • status: existing / generate / revise / audit
  3. Generate only the missing or weak figures first.
  4. Audit before handing the figure set to /paper-write or /paper-writing.

Practical Defaults

  • For plots, prefer the style habits adapted from MLNLP-World/Paper-Picture-Writing-Code, but normalized through templates/figure/matplotlib_publication_style.py.
  • For PGFPlots/TikZ, import templates/figure/pgfplots_bodyfont.tex and keep the figure text size equal to manuscript body text.
  • For architecture figures, prefer /figure-spec over ad hoc draw.io-style manual work unless the user explicitly asks for a manual editing workflow.

Output

Produce a compact decision record in the working notes or response:

Figure inventory:
- Fig 1: architecture overview -> /figure-spec -> generate
- Fig 2: main results bar chart -> /paper-figure -> generate
- Fig 3: ablation heatmap -> /paper-figure -> revise
- Table 1: benchmark comparison -> /paper-figure -> generate

Global rules:
- figure text size = paper body text size
- vector-first export
- final pass: /figure-table-audit

Completion Rule

Do not stop at "use skill X". Either:

  • invoke the downstream skill immediately, or
  • if the user asked only for process guidance, return the routed figure plan and the typography/export constraints.

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.