skills/paper-banana/SKILL.md
学术插图生成 - 使用 PaperBanana 多智能体框架从方法文本自动生成框架图和统计图
npx skillsauth add LiYu0524/Auto-Reasearch-Skills paper-bananaInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
使用 PaperBanana 多智能体框架(Planner → Visualizer → Critic 循环)从论文方法章节文本自动生成学术插图。
# 1. 克隆 PaperBanana
git clone https://github.com/paperbanana/PaperBanana.git ~/PaperBanana
# 2. 安装依赖
cd ~/PaperBanana
pip install -r requirements.txt
# 3. 配置模型(编辑 configs/model_config.yaml)
# 设置 OpenAI 兼容的 API base URL、API key 和模型名
SCRIPT=~/.claude/skills/auto-research/skills/paper-banana/scripts/generate_figure.py
# 基本用法
python3 $SCRIPT \
--content "方法文本(Markdown格式)" \
--caption "Figure 1: 框架图标题" \
--output ./figure.png
# 从文件读取内容
python3 $SCRIPT \
--content @method_section.md \
--caption "Figure 1: Pipeline overview" \
--output ./fig1.png
# 高质量模式(完整管线 + 更多迭代)
python3 $SCRIPT \
--content @method.md \
--caption "Figure 2: Architecture" \
--output ./fig2.png \
--exp-mode demo_full \
--critic-rounds 5
# 生成统计图
python3 $SCRIPT \
--content "实验结果数据..." \
--caption "Figure 3: Performance comparison" \
--output ./fig3.png \
--task plot
| 参数 | 默认值 | 说明 |
|------|--------|------|
| --content | (必填) | 方法文本,支持 @filepath 从文件读取 |
| --caption | (必填) | 图表标题 / 视觉意图描述 |
| --output | (必填) | 输出图片路径(.png / .jpg) |
| --task | diagram | 任务类型:diagram(框架图)或 plot(统计图) |
| --aspect-ratio | 16:9 | 宽高比 |
| --exp-mode | demo_planner_critic | 管线模式 |
| --retrieval-setting | none | 参考检索策略 |
| --critic-rounds | 3 | 最大 Critic 迭代轮数 |
| --image-model-name | (配置文件) | 覆盖图像生成模型 |
| --paperbanana-dir | ~/PaperBanana | PaperBanana 项目路径 |
| 模式 | 流程 | 适用场景 |
|------|------|----------|
| demo_planner_critic | Planner → Visualizer → Critic × N | 快速生成,推荐默认 |
| demo_full | Retriever → Planner → Stylist → Visualizer → Critic × N | 更精美,含风格优化 |
脚本输出 JSON 到 stdout:
{
"status": "success",
"output": "/absolute/path/to/figure.png",
"format": "PNG",
"size": "1920x1080",
"exp_mode": "demo_planner_critic",
"task": "diagram"
}
失败时:
{
"status": "error",
"message": "错误描述"
}
| 维度 | paper-banana | drawio | |------|-------------|--------| | 输出类型 | 位图(PNG/JPG),AI 生成 | 矢量图(XML),手动/规则生成 | | 适用场景 | 论文方法图、框架图、示意图 | 流程图、ER图、架构图 | | 可编辑性 | 不可编辑(位图) | 完全可编辑(XML) | | 视觉质量 | 高(AI风格化) | 中(工程图风格) | | 生成速度 | 慢(多轮 AI 调用) | 快(直接生成 XML) |
testing
Review research papers (especially PDFs). Use when the user asks to read/通读/讲解/总结/审稿 a paper and wants a Chinese-first explanation of what it does, what is novel (创新点), plus reviewer-style strengths/weaknesses, major/minor concerns, and questions to authors.
testing
End-to-end paper research support for arXiv/literature surveys, reproducibility-focused paper shortlisting, and experiment design. Use when you need to (1) search arXiv with complex queries, (2) download PDFs, extract text/sections, and fetch BibTeX, (3) dedupe/cluster results into a structured report, and (4) turn findings into a lit-review plan, benchmark/evaluation suite, and representation/probing experiment checklist (e.g., implicit reasoning, hidden-CoT, multilingual reasoning, cross-lingual alignment).
development
Manage Google Docs and Google Drive with full document operations and file management. Includes Markdown support for creating formatted documents with headings, bold, italic, lists, tables, and checkboxes. Also supports Drive operations (upload, download, share, search).
data-ai
Generate draw.io diagrams as .drawio files, optionally export to PNG/SVG/PDF with embedded XML