research-to-diagram

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Research to Diagram

深度调研主题并自动生成知识关系图谱PDF。从研究到可视化的一站式工具。

Conduct in-depth research on topics and automatically generate knowledge relationship graph PDFs. A one-stop tool from research to visualization.

Description

这个 Skill 接收用户提供的研究主题，自动进行深度网络调研、信息收集、知识整理、结构设计，最终生成专业的可视化关系图谱PDF。与

structure-to-pdf

不同，本 Skill 专注于主动研究和知识挖掘，而非被动的数据转换。

This Skill receives research topics provided by users, automatically performs in-depth web research, information collection, knowledge organization, structure design, and finally generates professional visualized relationship graph PDFs. Different from

structure-to-pdf

, this Skill focuses on active research and knowledge mining rather than passive data conversion.

核心特性

Core Features

自动调研：使用 WebSearch 进行多轮深度调研
智能整理：自动提取、分类、结构化信息
专业设计：根据主题特点选择最佳可视化方案
多种输出：支持 Graphviz、PlantUML、Mermaid 等工具
高质量PDF：生成矢量图形，可无限缩放

Automated Research: Conduct multi-round in-depth research using WebSearch
Intelligent Organization: Automatically extract, classify, and structure information
Professional Design: Select the optimal visualization scheme based on topic characteristics
Multiple Outputs: Support tools like Graphviz, PlantUML, Mermaid
High-Quality PDF: Generate vector graphics that can be infinitely scaled

Trigger Conditions

当用户想要：

研究某个复杂主题的知识结构（如"红楼梦人物关系"）
生成人物关系图、概念图谱、知识图谱
理解某个领域的组织架构、技术架构
可视化复杂的关系网络
需要从零开始研究并可视化某个主题

关键词：

"调研...并做图"
"研究...的关系"
"深度分析...并可视化"
"生成...知识图谱"
"画...关系图"（无现成数据）

When users want to:

Research the knowledge structure of a complex topic (e.g., "A Dream of Red Mansions character relationships")
Generate character relationship diagrams, concept maps, knowledge graphs
Understand the organizational structure or technical architecture of a certain field
Visualize complex relationship networks
Need to research and visualize a topic from scratch

Keywords:

"Research...and diagram"
"Study the relationships of..."
"In-depth analysis...and visualization"
"Generate...knowledge graph"
"Draw...relationship diagram" (no existing data)

Workflow

1. 任务规划（TodoWrite）

1. Task Planning (TodoWrite)

- 深度调研主题和相关知识
- 设计图谱结构和层次
- 创建可视化图表
- 生成PDF文档

- Conduct in-depth research on the topic and related knowledge
- Design the graph structure and hierarchy
- Create visual diagrams
- Generate PDF documents

2. 深度调研阶段

2. In-Depth Research Phase

多轮 WebSearch：从不同角度收集信息
- 主题概述和背景
- 核心要素和人物/概念
- 关系和联系
- 层次和分类
信息源记录：保存所有参考资料链接
知识提取：识别关键实体和关系

Multi-round WebSearch: Collect information from different perspectives
- Topic overview and background
- Core elements and characters/concepts
- Relationships and connections
- Hierarchies and classifications
Information Source Recording: Save links to all reference materials
Knowledge Extraction: Identify key entities and relationships

3. 结构设计阶段

3. Structure Design Phase

根据主题类型选择最佳结构：

人物关系图：

家族谱系：多层次树状结构
社会网络：网状关系图
组织架构：层次化布局

概念图谱：

知识分类：树状或思维导图
概念关系：有向图
流程图：线性或分支流程

技术架构：

系统组件：模块化布局
依赖关系：层次或网络图
数据流向：流程图

Select the optimal structure based on topic type:

Character Relationship Diagrams:

Family pedigree: Multi-level tree structure
Social network: Network relationship diagram
Organizational structure: Hierarchical layout

Concept Maps:

Knowledge classification: Tree or mind map
Concept relationships: Directed graph
Flowchart: Linear or branching process

Technical Architecture:

System components: Modular layout
Dependency relationships: Hierarchical or network diagram
Data flow: Flowchart

4. 可视化实现

4. Visualization Implementation

优先使用 Graphviz (DOT 语言)：

dot

digraph G {
    // 全局样式
    graph [rankdir=TB bgcolor="#fdfdf5" fontname="Arial Unicode MS"]
    node [shape=box style="rounded,filled" fillcolor="#e8f4f8"]
    edge [fontname="Arial Unicode MS"]

    // 使用 subgraph cluster 分组
    subgraph cluster_group1 {
        label="分组名称"
        node1 [label="节点1"]
        node2 [label="节点2"]
    }

    // 定义关系
    node1 -> node2 [label="关系类型" color=red]
}

备选工具：

PlantUML：UML 图、时序图
Mermaid：简单流程图、时序图

Priority use of Graphviz (DOT language):

dot

digraph G {
    // 全局样式
    graph [rankdir=TB bgcolor="#fdfdf5" fontname="Arial Unicode MS"]
    node [shape=box style="rounded,filled" fillcolor="#e8f4f8"]
    edge [fontname="Arial Unicode MS"]

    // 使用 subgraph cluster 分组
    subgraph cluster_group1 {
        label="分组名称"
        node1 [label="节点1"]
        node2 [label="节点2"]
    }

    // 定义关系
    node1 -> node2 [label="关系类型" color=red]
}

Alternative Tools:

PlantUML: UML diagrams, sequence diagrams
Mermaid: Simple flowcharts, sequence diagrams

5. PDF 生成

5. PDF Generation

bash

dot -Tpdf diagram.dot -o output.pdf

bash

dot -Tpdf diagram.dot -o output.pdf

6. 文档整理

6. Document Organization

可选生成说明文档，包含：

研究主题概述
图谱说明
参考资料来源（Sources）
使用说明

Optionally generate explanatory documents including:

Research topic overview
Graph explanation
Reference sources (Sources)
Usage instructions

设计原则

Design Principles

视觉设计

Visual Design

颜色编码：使用不同颜色区分类别
形状区分：不同类型实体用不同形状
层次清晰：使用 subgraph cluster 分组
关系标注：边的颜色、样式、标签表达关系类型
中文支持：使用 "Arial Unicode MS" 或系统中文字体

Color Coding: Use different colors to distinguish categories
Shape Differentiation: Use different shapes for different types of entities
Clear Hierarchy: Use subgraph cluster for grouping
Relationship Annotation: Use edge color, style, and labels to express relationship types
Chinese Support: Use "Arial Unicode MS" or system Chinese fonts

信息层次

Information Hierarchy

标题层：主标题
分组层：主要类别/家族
实体层：具体人物/概念
关系层：连接和标注
图例层：说明符号含义

Title Layer: Main title
Grouping Layer: Main categories/families
Entity Layer: Specific characters/concepts
Relationship Layer: Connections and annotations
Legend Layer: Explain symbol meanings

布局策略

Layout Strategies

rankdir=TB：自上而下（家族树、组织架构）
rankdir=LR：从左到右（流程图、时间线）
rankdir=BT：自下而上（依赖图）
splines=ortho：正交边（清晰的组织图）
splines=curved：曲线边（美观的关系网）

rankdir=TB: Top to bottom (family trees, organizational structures)
rankdir=LR: Left to right (flowcharts, timelines)
rankdir=BT: Bottom to top (dependency diagrams)
splines=ortho: Orthogonal edges (clear organizational diagrams)
splines=curved: Curved edges (aesthetic relationship networks)

输出文件

Output Files

默认保存位置：

~/Downloads/

或用户指定目录

生成文件：

```
<topic>_relations.dot
```
- Graphviz 源文件
```
<topic>_relations.pdf
```
- 最终PDF图谱
```
<topic>_sources.md
```
- 参考资料（可选）

Default save location:

~/Downloads/

or user-specified directory

Generated files:

```
<topic>_relations.dot
```
- Graphviz source file
```
<topic>_relations.pdf
```
- Final PDF graph
```
<topic>_sources.md
```
- Reference materials (optional)

使用示例

Usage Examples

示例 1：文学作品人物关系

Example 1: Literary Work Character Relationships

用户：深度调查《三国演义》里人物之间的关系，然后做个结构图 PDF

Skill 执行：

调研三国主要人物、阵营、关系
设计：魏蜀吴三大阵营 + 人物层次 + 联盟/对抗关系
使用 Graphviz 创建多层次关系图
生成 PDF

User: Conduct in-depth research on the relationships between characters in "Romance of the Three Kingdoms" and create a structure diagram PDF

Skill Execution:

Research main characters, camps, and relationships in the Three Kingdoms period
Design: Three major camps (Wei, Shu, Wu) + character hierarchy + alliance/opposition relationships
Create a multi-level relationship diagram using Graphviz
Generate PDF

示例 2：技术概念图谱

Example 2: Technical Concept Map

用户：研究 Kubernetes 架构并生成可视化图谱

Skill 执行：

调研 K8s 核心组件、架构层次
设计：控制平面/数据平面/插件生态
生成技术架构图
输出 PDF

User: Research Kubernetes architecture and generate a visualized graph

Skill Execution:

Research K8s core components and architecture hierarchy
Design: Control plane/data plane/plugin ecosystem
Generate technical architecture diagram
Output PDF

示例 3：历史事件关系

Example 3: Historical Event Relationships

用户：分析二战主要国家和联盟关系，做成图表

Skill 执行：

调研参战国家、阵营、关键时间节点
设计：轴心国/同盟国/中立国关系网
生成带时间线的关系图
输出 PDF

User: Analyze the relationships between major countries and alliances in World War II and create a chart

Skill Execution:

Research participating countries, camps, and key time nodes
Design: Axis powers/Allied powers/neutral countries relationship network
Generate a relationship diagram with timeline
Output PDF

与 structure-to-pdf 的区别

Differences from structure-to-pdf

特性	research-to-diagram	structure-to-pdf
输入	仅主题/研究问题	现成的结构化数据
调研	✅ 自动深度调研	❌ 无需调研
知识整理	✅ 自动提取和结构化	❌ 直接使用用户数据
应用场景	知识探索、研究可视化	快速数据转换
时间	较长（需调研）	快速

何时使用本 Skill：

✅ 没有现成数据，需要从零研究
✅ 想要深度了解某个主题的知识结构
✅ 需要权威来源支持的可视化

何时使用 structure-to-pdf：

✅ 已有结构化数据
✅ 需要快速转换为图表
✅ 数据格式简单清晰

Feature	research-to-diagram	structure-to-pdf
Input	Only topic/research question	Existing structured data
Research	✅ Automated in-depth research	❌ No research needed
Knowledge Organization	✅ Automatic extraction and structuring	❌ Directly use user data
Application Scenarios	Knowledge exploration, research visualization	Quick data conversion
Time	Longer (requires research)	Fast

When to Use This Skill:

✅ No existing data, need to research from scratch
✅ Want to deeply understand the knowledge structure of a topic
✅ Need visualization supported by authoritative sources

When to Use structure-to-pdf:

✅ Have existing structured data
✅ Need quick conversion to charts
✅ Data format is simple and clear

依赖工具

Dependent Tools

必需：

Graphviz:
```
brew install graphviz
```
(macOS)
WebSearch: Claude Code 内置

可选：

PlantUML:
```
brew install plantuml
```
Mermaid CLI:
```
npm install -g @mermaid-js/mermaid-cli
```

Required:

Graphviz:
```
brew install graphviz
```
(macOS)
WebSearch: Built into Claude Code

Optional:

PlantUML:
```
brew install plantuml
```
Mermaid CLI:
```
npm install -g @mermaid-js/mermaid-cli
```

配置选项

Configuration Options

用户可以通过参数自定义：

```
--tool
```
: 指定可视化工具 (graphviz/plantuml/mermaid)
```
--layout
```
: 指定布局方向 (TB/LR/BT/RL)
```
--output
```
: 指定输出目录
```
--depth
```
: 调研深度 (quick/medium/deep)
```
--sources
```
: 是否生成参考资料文档

Users can customize via parameters:

```
--tool
```
: Specify visualization tool (graphviz/plantuml/mermaid)
```
--layout
```
: Specify layout direction (TB/LR/BT/RL)
```
--output
```
: Specify output directory
```
--depth
```
: Research depth (quick/medium/deep)
```
--sources
```
: Whether to generate reference materials document

最佳实践

Best Practices

明确主题范围：主题越具体，图谱越清晰
合理分组：使用 cluster 将相关实体分组
控制复杂度：单个图谱不超过 50 个节点
渐进细化：先生成总览图，再深入细节
颜色一致：同类实体使用相同配色
标注来源：在 PDF 底部或单独文档标注参考资料

Clear Topic Scope: The more specific the topic, the clearer the graph
Reasonable Grouping: Use clusters to group related entities
Control Complexity: No more than 50 nodes in a single graph
Progressive Refinement: Generate an overview first, then dive into details
Consistent Colors: Use the same color scheme for similar entities
Cite Sources: Label reference materials at the bottom of the PDF or in a separate document

常见图谱类型模板

Common Graph Type Templates

人物关系图模板

Character Relationship Diagram Template

dot

- 家族/组织用 cluster 分组
- 人物用 box/ellipse，重要人物用特殊形状
- 血缘关系用实线，婚姻用红色，其他用虚线
- 添加图例说明符号含义

dot

- Use cluster for family/organization grouping
- Use box/ellipse for characters, special shapes for important characters
- Use solid lines for blood relationships, red lines for marriage, dashed lines for others
- Add legends to explain symbol meanings

概念图谱模板

Concept Map Template

dot

- 顶层概念在上方
- 子概念逐层展开
- is-a 关系用实线，has-a 用虚线
- 使用颜色区分不同类别

dot

- Top-level concepts at the top
- Sub-concepts expanded layer by layer
- Use solid lines for is-a relationships, dashed lines for has-a relationships
- Use colors to distinguish different categories

技术架构图模板

Technical Architecture Diagram Template

dot

- 分层架构用 rankdir=TB
- 组件用矩形，服务用圆角矩形
- 依赖关系用箭头
- 关键路径用粗线或特殊颜色

dot

- Use rankdir=TB for layered architecture
- Use rectangles for components, rounded rectangles for services
- Use arrows for dependency relationships
- Use thick lines or special colors for critical paths

故障排除

Troubleshooting

中文显示乱码：

dot

graph [fontname="Arial Unicode MS"]
node [fontname="Arial Unicode MS"]
edge [fontname="Arial Unicode MS"]

图谱过于复杂：

使用
```
concentrate=true
```
合并相同边
分割成多个子图
使用不同的 rankdir

布局不理想：

调整
```
ranksep
```
和
```
nodesep
```
使用
```
rank=same
```
强制节点同层
尝试不同的
```
splines
```
设置

Chinese Display Garbled Characters:

dot

graph [fontname="Arial Unicode MS"]
node [fontname="Arial Unicode MS"]
edge [fontname="Arial Unicode MS"]

Overly Complex Graph:

Use
```
concentrate=true
```
to merge identical edges
Split into multiple subgraphs
Use different rankdir settings

Ideal Layout Not Achieved:

Adjust
```
ranksep
```
and
```
nodesep
```
Use
```
rank=same
```
to force nodes to be on the same layer
Try different
```
splines
```
settings

版本历史

Version History

v1.0 (2026-01-02): 初始版本
- 基于《红楼梦》人物关系图谱项目总结
- 支持 Graphviz 自动生成
- 集成 WebSearch 深度调研
- TodoWrite 任务管理

v1.0 (2026-01-02): Initial version
- Summarized based on the "A Dream of Red Mansions" character relationship graph project
- Support automatic generation with Graphviz
- Integrated WebSearch for in-depth research
- TodoWrite for task management