/ 目录 / 演练场 / paperbanana-skill
● 社区 PlutoLei ⚡ 即开即用

paperbanana-skill

作者 PlutoLei · PlutoLei/paperbanana-skill

Generate publication-quality academic diagrams — architecture figures, training curves, system schematics — from prompts, PaperBanana-style.

A Claude Code skill wrapping the PaperBanana workflow for academic figures. Instead of fighting TikZ or Inkscape, describe the figure; get publishable SVG/PDF output with the visual grammar papers expect (clean typography, consistent arrows, no rainbow colors).

为什么要用

核心特性

实时演示

实际使用效果

paperbanana-skill.replay ▶ 就绪
0/0

安装

作者仓库 · Claude Skills 官方文档

使用场景

实战用法: paperbanana-skill

Draw your model architecture for a conference paper

👤 ML researchers, grad students mid-paper ⏱ ~30 min intermediate

何时使用: Deadline in 5 days, you still don't have a clean architecture figure.

步骤
  1. Describe the model
    Use paperbanana-skill. Architecture figure: input → tokenizer → transformer stack (6 layers) → classifier head. Use the skill's default paper style.✓ 已复制
    → SVG with clean boxes, arrows, consistent typography
  2. Tweak
    Label the transformer block internals — attention + FFN.✓ 已复制
    → Updated figure with sub-block detail
  3. Export for LaTeX
    Export as PDF sized for a two-column paper.✓ 已复制
    → PDF with correct aspect ratio

结果: A reviewer-worthy figure without a week of Inkscape.

注意事项
  • Figure over-decorated — Prefer restraint; one accent color, black arrows, lots of whitespace
搭配使用: filesystem

Turn training logs into a publication-ready plot

👤 ML researchers with wandb/tensorboard exports ⏱ ~20 min beginner

何时使用: You have CSVs of training metrics and need a figure for the paper.

步骤
  1. Feed CSV
    Use paperbanana-skill. Plot loss and accuracy curves from training.csv in two-panel layout.✓ 已复制
    → Two-panel figure with clean axes
  2. Add baselines
    Overlay the baseline model's curves (baseline.csv) as dashed.✓ 已复制
    → Overlay added with clear legend

结果: A plot that fits the paper's aesthetic.

Keep all figures in a paper visually consistent

👤 Authors with 6+ figures drifting in style ⏱ ~45 min intermediate

何时使用: You look at the paper and realize each figure has a different font.

步骤
  1. Define the style
    Use paperbanana-skill. Create a shared style: 10pt sans-serif, #1f77b4 primary, black stroke 0.8pt.✓ 已复制
    → Style file saved
  2. Apply to all figures
    Regenerate figures 1-6 using this style.✓ 已复制
    → Consistent batch output

结果: Figures that look like they belong in the same paper.

搭配使用: filesystem

组合

与其他 MCP 搭配,撬动十倍杠杆

paperbanana-skill + filesystem

Organize figures in the paper's /figures directory

Save all figures under paper/figures/ with source.svg + out.pdf per figure.✓ 已复制
paperbanana-skill + mermaid-skill

Use Mermaid for casual diagrams, PaperBanana for publication quality

Sketch in Mermaid for discussion; produce with PaperBanana for camera-ready.✓ 已复制

工具

此 MCP 暴露的能力

工具输入参数何时调用成本
architecture_diagram description System figures 0
plot_curves data file, metrics Training/eval metrics 0
apply_style style file, figures Cross-figure consistency 0

成本与限制

运行它的成本

API 配额
None
每次调用 Token 数
Moderate
费用
Free
提示
Lock the style once; apply across figures to avoid re-deciding.

安全

权限、密钥、影响范围

凭据存储: None
数据出站: None from skill

故障排查

常见错误与修复

Exported PDF has wrong font in LaTeX

Embed fonts or use standard PDF font families; check paperbanana's export settings

SVG renders differently in browser vs LaTeX

Stick to PDF for LaTeX; SVG for web/slides

替代方案

paperbanana-skill 对比其他方案

替代方案何时用它替代权衡
TikZ / PGFPlotsYou want full LaTeX integration and are comfortable with TikZSteep learning curve
mermaid-skillCasual docs, not publicationNot publication-quality aesthetic

更多

资源

📖 阅读 GitHub 上的官方 README

🐙 查看未解决的 issue

🔍 浏览全部 400+ MCP 服务器和 Skills