S2 multi-dimensional cross-analysis table development assistant. Use when users need help with S2 pivot tables, table sheets, or any @antv/s2 related development.

GitHub stars
536

Install

/plugin install antv-s2-expert@visualization-skills

Run it inside Claude Code after adding the marketplace with /plugin marketplace add antvis/chart-visualization-skills.

Contents

Discovered from source

Skills 9

G2 v5 Chart Visualization

Overview

G2 v5 is AntV's grammar-of-graphics charting library. It uses Spec Mode — a declarative, JSON-like configuration style. Generated examples should start from a complete spec; runtime code may later merge a local update through chart.options().

import { Chart } from '@antv/g2';

const chart = new Chart({ container: 'container', autoFit: true });
chart.options({
  type: 'interval',
  data: [{ genre: 'Sports', sold: 275 }],
  encode: { x: 'genre', y: 'sold' },
});
chart.render();

CDN Usage

<script src="https://unpkg.com/@antv/g2@5/dist/g2.min.js"></script>
<script>
  const chart = new G2.Chart({ container: 'container', autoFit: true });
  chart.options({
    type: 'interval',
    data: [{ genre: 'Sports', sold: 275 }],
    encode: { x: 'genre', y: 'sold' },
  });
  chart.render();
</script>

Content Retrieval Service

When using AntV G2 for data visualization, if you need to understand the concepts, usage, API, examples, and other aspects of G2 v5, you can use the provided context retrieval service. When using the skill, content is retrieved via an antv HTTP API server using GET requests.

  • Host: https://sive.antv.antgroup.com
  • Endpoint: /api/v1/context/retrieve
  • Method: GET
  • Parameters: query, library, topK, content, maxTokens

Retrieve reference documents by query (hybrid search = FTS + vector + RRF fusion).

ParameterTypeRequiredDescription
querystring✅Search keywords, e.g. bar chart interval
librarystring✅Library name: g2, g6, x6
topKnumberNumber of results to return (default: 5)
contentbooleanReturn full reference doc markdown (default: true)
maxTokensnumberMax tokens per result (default: unlimited)
curl "https://sive.antv.antgroup.com/api/v1/context/retrieve?query=bar+chart+stacked&library=g2"

Critical Rules

MUST: Use V5 Spec Mode ONLY

// ❌ WRONG — V4 chain API (deprecated, will not render)
chart.interval()
  .data([...])
  .encode('x', 'genre')
  .encode('y', 'sold')
  .style({ radius: 4 });

// ✅ CORRECT — V5 Spec Mode
chart.options({
  type: 'interval',
  data: [...],
  encode: { x: 'genre', y: 'sold' },
  style: { radius: 4 },
});

Multi-mark overlays use view + children

chart.options() supports incremental deep-merge updates, which is useful for runtime interaction. For generated one-shot code, provide a complete initial spec. To create independent overlay marks, use one type: 'view' with children; sequentially changing the root type does not create an overlay.

// ❌ WRONG — this changes one root mark from line to point; it does not overlay them
chart.options({ type: 'line', data, encode: { x: 'date', y: 'value' } });
chart.options({ type: 'point', data, encode: { x: 'date', y: 'value' } });

// ✅ CORRECT — children array for multi-mark
chart.options({
  type: 'view',
  data,
  children: [
    { type: 'line',  encode: { x: 'date', y: 'value' } },
    { type: 'point', encode: { x: 'date', y: 'value' } },
  ],
});

MUST: container is mandatory, chart.render() at the end

// ❌ WRONG — no container, no render
const chart = new Chart();
chart.options({ type: 'interval', data });

// ✅ CORRECT
const chart = new Chart({ container: 'container', autoFit: true });
chart.options({ type: 'interval', data, encode: { x: 'genre', y: 'sold' } });
chart.render();

MUST: Correct mark types only

❌ Hallucinated (from ECharts/Vega)✅ G2 correct replacement
type: 'ruleX'type: 'lineX'
type: 'ruleY'type: 'lineY'
type: 'regionX'type: 'rangeX'
type: 'regionY'type: 'rangeY'
type: 'venn'type: 'path' + transform

Legal G2 marks: interval, line, area, point, rect, cell, text, image, path, polygon, shape, link, connector, vector, lineX, lineY, rangeX, rangeY, range, box, boxplot, density, heatmap, beeswarm, treemap, pack, partition, tree, sankey, chord, wordCloud, gauge, liquid. sunburst requires @antv/g2-extension-plot.

MUST: encode is an object, transform is an array

// ❌ WRONG
.encode('x', 'genre')
.transform: { type: 'stackY' }

// ✅ CORRECT
encode: { x: 'genre', y: 'sold' }
transform: [{ type: 'stackY' }]

MUST: labels is plural; range encodings use mark-appropriate channels

// ❌ WRONG
label: { text: 'sold' }

// ✅ CORRECT
labels: [{ text: 'sold' }]
// Interval / link ranges may use either a two-field array or y + y1.
encode: { y: ['start', 'end'] }
// RangeY explicitly uses y + y1.
encode: { y: 'start', y1: 'end' }

MUST: No d3 in user code

// ❌ WRONG — d3 is not exposed in user scope
const total = d3.sum(data, d => d.value);

// ✅ CORRECT — use native JS or G2 built-in transforms
const total = data.reduce((sum, d) => sum + d.value, 0);

MUST: No white/near-white fill, no padding as array

// ❌ WRONG
style: { fill: '#fff' }       // invisible on white background
padding: [40, 30, 40, 50]     // invalid in G2 v5

// ✅ CORRECT
encode: { color: 'group' }    // let G2 assign colors
padding: 40                   // single number or 'auto'

MUST: Transpose is a transform, not a coordinate type

// ❌ WRONG
coordinate: { type: 'transpose' }

// ✅ CORRECT
coordinate: { transform: [{ type: 'transpose' }] }

Default Aesthetics / 默认视觉基线

目标是默认清晰、克制、可读,而不是给每张图添加相同装饰。先遵守以下决策,再按需检索 default+aesthetics+design 获取完整示例。

注意区分两种“默认”。G2 引擎默认(theme/create.ts)是 line.lineWidth: 1、area.fillOpacity: 0.85、矩形 radius 无圆角;本节 Mark 基线里的 radius: 4 / lineWidth: 2 / fillOpacity: 0.5~0.6 是本 skill 的约定值,用来覆写引擎默认让生成示例更好看,必须显式写在 style 里,不是省略即生效。

  1. 通用:container、完整初始 chart.options() spec、chart.render() 是必须项;普通嵌入式图优先 autoFit: true、theme: 'classic' 与 padding: 'auto'。classic 是为稳定视觉显式选择的浅色预设,不是引擎默认主题(引擎默认推断为 light);深色容器使用 classicDark。两者共用同一套 category10 色板(首选 #5B8FF9),因此稳定主色在浅/深主题下都可用。
  2. 颜色表达语义:仅当颜色表示独立的系列、状态或分组时使用 encode.color 和图例。单指标分类比较使用稳定单色(如 style.fill: '#5B8FF9',作为样式值是合法的)并关闭颜色图例,避免彩虹柱和冗余图例。禁止的是把 hex 字符串存进数据后作为 encode.color 的类别字段编码。
  3. Mark 基线(约定值,覆写引擎默认):interval 用 radius: 4(引擎默认无圆角);line 用 lineWidth: 2(引擎默认 1);area 用 fillOpacity: 0.5~0.6(引擎默认 0.85,偏实,降下来更透气)。这些值都需显式写在 style 里,不是省略即生效;不需要渐变、阴影或自定义动画。
  4. 文本只用已知语义:字段/单位/来源明确时,添加语义化轴标题、tooltip 名称和 formatter;报告语境或用户提供标题时添加顶层 title。信息未知时省略,不编造单位、来源或副标题。
  5. 标签按密度选择:少量且需精确读取的数据可加标签;密集散点、多系列折线、类别很多时默认依赖 tooltip。inside 标签用 contrastReverse,发生碰撞时使用 overlapHide / overlapDodgeY / overflowHide;dx / dy 只用于避让后的细微调整。
  6. 特殊图:饼/环图的少量类别优先外置标签,类别较多时改用 legend;不要默认同时重复两者。气泡图保持 G2 默认 sqrt size 映射,按数据范围设置 size.range,仅在 size legend 无助理解时隐藏它。

渐变、阴影、滑块、滚动条、自定义动画和 3D 气泡都属于用户明确要求“报告级 / 精致”时的增强项;添加前必须确认不会掩盖数据或降低标签对比度。

Quick Reference

User IntentRetrieve Query
Chart initialization, container, autoFitGET /api/v1/context/retrieve?query=chart+init&library=g2
Bar / column chartGET /api/v1/context/retrieve?query=bar+chart+interval&library=g2
Line / area chartGET /api/v1/context/retrieve?query=line+area+chart&library=g2
Pie / donut / rose chartGET /api/v1/context/retrieve?query=pie+chart+theta&library=g2
Scatter / bubbleGET /api/v1/context/retrieve?query=scatter+point+bubble&library=g2
Treemap / sunburst / packGET /api/v1/context/retrieve?query=treemap+sunburst+pack&library=g2
Heatmap / density / boxplotGET /api/v1/context/retrieve?query=heatmap+density+boxplot&library=g2
Funnel / gauge / wordcloudGET /api/v1/context/retrieve?query=funnel+gauge+wordcloud&library=g2
Encode channels (x, y, color, size)GET /api/v1/context/retrieve?query=encode+channel&library=g2
Scale / palette / color rangeGET /api/v1/context/retrieve?query=scale+palette+color&library=g2
Coordinate (polar, theta, transpose)GET /api/v1/context/retrieve?query=coordinate+polar+theta+transpose&library=g2
Transform (stack, normalize, sort)GET /api/v1/context/retrieve?query=transform+stack+normalize&library=g2
Axis / legend / tooltip / labelsGET /api/v1/context/retrieve?query=axis+legend+tooltip+label&library=g2
Interaction (brush, highlight, drilldown)GET /api/v1/context/retrieve?query=interaction+brush+highlight&library=g2
Theme / dark modeGET /api/v1/context/retrieve?query=theme+dark+classicDark&library=g2
Default aesthetics / 视觉基线GET /api/v1/context/retrieve?query=default+aesthetics+design&library=g2
AnimationGET /api/v1/context/retrieve?query=animation+animate&library=g2
Data fetch / filter / sortGET /api/v1/context/retrieve?query=data+fetch+filter+sort&library=g2
Facet / view compositionGET /api/v1/context/retrieve?query=facet+view+composition&library=g2
Chart type selection guideGET /api/v1/context/retrieve?query=chart+type+selection&library=g2
Rendering troubleshootGET /api/v1/context/retrieve?query=rendering+troubleshoot+debug&library=g2

Dependencies

  • @antv/g2 — G2 v5 charting engine

G6 v5 Graph Visualization

Overview

G6 v5 is AntV's graph visualization engine for network diagrams, tree graphs, and relationship visualizations. It uses a declarative configuration style where new Graph({...}) defines all nodes, edges, layouts, behaviors, and plugins in one constructor call.

import { Graph } from '@antv/g6';

const graph = new Graph({
  container: 'container',
  data: {
    nodes: [{ id: 'node-1', style: { labelText: 'Node 1' } }],
    edges: [{ source: 'node-1', target: 'node-2' }],
  },
  layout: { type: 'force' },
  behaviors: ['drag-canvas', 'zoom-canvas', 'drag-element'],
});

await graph.render();

CDN Usage

<script src="https://unpkg.com/@antv/g6@5/dist/g6.min.js"></script>
<script>
  const graph = new G6.Graph({
    container: 'container',
    data: {
      nodes: [{ id: 'node-1', style: { labelText: 'Node 1' } }],
      edges: [{ source: 'node-1', target: 'node-2' }],
    },
    layout: { type: 'force' },
    behaviors: ['drag-canvas', 'zoom-canvas', 'drag-element'],
  });
  graph.render();
</script>

Content Retrieval Service

When using AntV G6 for data visualization, if you need to understand the concepts, usage, API, examples, and other aspects of G6 v5, you can use the provided context retrieval service. When using the skill, content is retrieved via an antv HTTP API server using GET requests.

  • Host: https://sive.antv.antgroup.com
  • Endpoint: /api/v1/context/retrieve
  • Method: GET
  • Parameters: query, library, topK, content, maxTokens

Retrieve skills by query (hybrid search = FTS + vector + RRF fusion). Constraints docs are indexed as regular skill documents and will appear in search results naturally.

ParameterTypeRequiredDescription
querystring✅Search keywords, e.g. force layout node
librarystring✅Library name: g2, g6, x6
topKnumberNumber of results to return (default: 5)
contentbooleanReturn full reference doc markdown (default: true)
maxTokensnumberMax tokens per result (default: unlimited)
curl "https://sive.antv.antgroup.com/api/v1/context/retrieve?query=force+layout+node+style&library=g6"

Critical Rules

MUST: Use new Graph({...}) — NOT v4 new G6.Graph()

// ❌ WRONG — v4 constructor
new G6.Graph({ container: 'container', ... });

// ✅ CORRECT — v5 constructor
import { Graph } from '@antv/g6';
new Graph({ container: 'container', ... });

MUST: All config in one constructor call, await graph.render()

// ❌ WRONG — v4 separate data method
graph.data(data);
graph.render();

// ✅ CORRECT — v5 declarative config + async render
const graph = new Graph({
  container: 'container',
  data: { nodes: [...], edges: [...] },
  layout: { type: 'force' },
  behaviors: ['drag-canvas', 'zoom-canvas'],
});
await graph.render();

MUST: Data format with id, source, target

// ❌ WRONG — missing node id, missing edge endpoints
const data = { nodes: [{ label: 'A' }], edges: [{ from: 'A', to: 'B' }] };

// ✅ CORRECT — each node has unique id, each edge has source/target
const data = {
  nodes: [{ id: 'node-1', style: { labelText: 'A' } }],
  edges: [{ source: 'node-1', target: 'node-2' }],
};

MUST: Use style.labelText for labels — NOT label or labelCfg

// ❌ WRONG — v4 label config
node: { labelCfg: { text: 'Node 1' } }

// ✅ CORRECT — v5 style.labelText
node: { style: { labelText: 'Node 1' } }

MUST: nodeStrength must be ≥ 0 in force layout

// ❌ WRONG — negative nodeStrength causes unpredictable behavior
layout: { type: 'force', nodeStrength: -300 }

// ✅ CORRECT — non-negative value
layout: { type: 'force', nodeStrength: 300 }

MUST: force layout does NOT support preventOverlap / nodeSize

// ❌ WRONG — v4 params silently ignored in v5
layout: { type: 'force', preventOverlap: true, nodeSize: 30 }

// ✅ CORRECT — use d3-force collide for overlap prevention
layout: { type: 'd3-force', collide: { radius: 30 } }

MUST: No Mode concept — behaviors are flat array

// ❌ WRONG — v4 mode-based behavior config
modes: { default: ['drag-canvas', 'zoom-canvas'] }

// ✅ CORRECT — v5 flat behavior array
behaviors: ['drag-canvas', 'zoom-canvas', 'drag-element']

MUST: container is mandatory, default 'container'

// ❌ WRONG — no container specified
const graph = new Graph({ data });

// ✅ CORRECT
const graph = new Graph({ container: 'container', data, ... });

Quick Reference

User IntentRetrieve Query
Graph initialization, container, renderGET /api/v1/context/retrieve?query=graph+init+render&library=g6
Network / force graphGET /api/v1/context/retrieve?query=network+force+layout&library=g6
Tree / mindmap / fishboneGET /api/v1/context/retrieve?query=tree+mindmap+fishbone+layout&library=g6
Dagre / hierarchy / flow chartGET /api/v1/context/retrieve?query=dagre+hierarchy+flow+chart&library=g6
Circular / radial / grid layoutGET /api/v1/context/retrieve?query=circular+radial+grid+layout&library=g6
Node styles (rect, circle, diamond, html)GET /api/v1/context/retrieve?query=node+style+rect+circle+diamond+html&library=g6
Edge types (line, cubic, polyline, loop)GET /api/v1/context/retrieve?query=edge+line+cubic+polyline+loop&library=g6
Combo / group nodesGET /api/v1/context/retrieve?query=combo+group+node&library=g6
Custom node / edgeGET /api/v1/context/retrieve?query=custom+node+edge+element&library=g6
Behaviors (drag, zoom, click-select, hover)GET /api/v1/context/retrieve?query=behavior+drag+zoom+click-select+hover&library=g6
Plugins (minimap, tooltip, toolbar, legend)GET /api/v1/context/retrieve?query=plugin+minimap+tooltip+toolbar+legend&library=g6
Events systemGET /api/v1/context/retrieve?query=events+system+click+mouse&library=g6
State / style animationGET /api/v1/context/retrieve?query=state+animation+transform&library=g6
Data structure / transformsGET /api/v1/context/retrieve?query=data+structure+transforms&library=g6
Theme / backgroundGET /api/v1/context/retrieve?query=theme+background+style&library=g6
Lasso select / collapse-expandGET /api/v1/context/retrieve?query=lasso+collapse+expand+select&library=g6

Dependencies

  • @antv/g6 — G6 v5 graph visualization engine

GPT-Vis Chart Visualization Skill

GPT-Vis is an AI-native visualization library designed for the LLM era. It uses a framework-agnostic architecture, supports 26 chart types, and enables LLMs to easily generate high-quality visualizations through a simple, natural syntax.

Steps

  1. Intent recognition: Select the chart type based on user intent and data characteristics
  2. Determine output mode: Choose Syntax mode or Code mode based on context
  3. Generate output: Produce content in the selected mode

Supported Chart Types

type valueUse case
lineTime series trends
areaTime series trends + totals
columnCategorical data comparison
barCategorical comparison (long labels)
piePart-to-whole proportions
scatterRelationship between two variables
dual-axesComparison of data at different scales
histogramFrequency distribution of continuous values
boxplotData distribution and outliers
violinData distribution density
radarMulti-dimensional comparison
funnelProcess conversion rates
waterfallCumulative increases and decreases
liquidPercentage / progress
word-cloudWord frequency display
vennSet intersection/union relationships
treemapHierarchical data proportions
sankeyFlow and direction of quantities
flow-diagramProcess steps
mindmapHierarchical knowledge organization
indented-treeTree node hierarchy / directories
network-graphRelationships between entities
organization-chartOrganizational hierarchy
fishbone-diagramRoot cause analysis
tableTabular data display
summaryContent summary

Output Modes

Mode 1: Syntax Mode (Syntax / JSON)

Used for LLM application integration scenarios. Generates chart configurations consumed by GPTVis.render(). Supports two formats:

  • Syntax format: Markdown-like indentation syntax, suitable for streaming output (renders in real time as the LLM generates tokens)
  • JSON format: Standard JSON object, suitable for structured API calls

Both formats are equivalent and can be passed directly to GPTVis.render().

Mode 2: Code Mode

Used when the user needs complete runnable code. Generates output including installation instructions and full code.

GPTVis API

GPTVis is the unified entry class of the library, responsible for creating, rendering, and destroying charts.

Constructor

new GPTVis(options: VisualizationOptions)

VisualizationOptions:

ParameterTypeRequiredDefaultDescription
containerstring | HTMLElementYes—CSS selector or DOM element
widthnumberNo—Chart width (px)
heightnumberNo—Chart height (px)
theme'default' | 'light' | 'dark' | 'academy'No'light'Theme
wrapperbooleanNofalseWhether to show the outer UI container (tabs, download, copy)
localestringNo'zh-CN'Language of the text inside the wrapper

Methods

render(config: string | object): void

Renders a chart. Accepts two kinds of input:

  • Syntax string: Text starting with vis [type], automatically parsed into a configuration object
  • JSON configuration object: An object containing a type field
  • Plain text: Strings not starting with vis are rendered as the summary type

Calling render() multiple times automatically destroys the previous chart before rendering the new one.

destroy(): void

Destroys the current chart instance and releases resources.

Syntax Mode: JSON Format

Directly output a JSON object conforming to the chart's TypeScript type; GPTVis.render() can consume it directly.

JSON Example

{
  "type": "column",
  "data": [
    { "category": "Product A", "value": 30, "group": "Online" },
    { "category": "Product B", "value": 50, "group": "Online" }
  ],
  "title": "Product Sales Comparison",
  "axisXTitle": "Product",
  "axisYTitle": "Sales (10k)",
  "stack": true,
  "theme": "academy",
  "style": {
    "palette": ["#5B8FF9", "#61DDAA"]
  }
}

Syntax Mode: Syntax Format

Markdown-like indentation syntax that supports streaming rendering. The first line must be vis [type].

Syntax Rules

Basic properties — key value, one per line:

title Annual Trend
theme dark

Object arrays — under data, each item starts with - , with sub-fields indented:

{ data: { time: string; value: number; }[]; }

Corresponds to:

data
  - time 2020
    value 100
  - time 2021
    value 120

Plain value arrays — each item starts with - :

{ data: number[] }

Corresponds to:

data
  - 10
  - 20

String values containing spaces — wrap in quotes (single or double); quotes can be omitted when there are no spaces:

categories
  - "North America"
  - 'Southeast Asia'
  - Europe

Nested objects — the object name occupies its own line, with sub-properties indented:

{ style?: { backgroundColor?: string; palette?: string[] } }

Corresponds to:

style
  backgroundColor #f0f2f5
  palette
    - #5B8FF9
    - #61DDAA

Recursive trees — the children array is indented with - :

type TreeData = { name: string; children?: TreeData[] };
{ data: TreeData; }

Corresponds to:

data
  name Root
  children
    - name Child A
      children
        - name Grandchild
    - name Child B

Complete Syntax Example

vis column
data
  - category Product A
    value 30
    group Online
  - category Product B
    value 50
    group Online
title Product Sales Comparison
axisXTitle Product
axisYTitle Sales (10k)
stack true
theme academy
style
  palette
    - #5B8FF9
    - #61DDAA

Markdown Syntax

When the output is in Markdown format, use GPT-Vis as the language tag of the fenced code block, and write the complete Syntax format in the content area:

```GPT-Vis
vis line
data
  - time 2020
    value 100
```

Format:

```GPT-Vis
<complete Syntax content, first line: vis <chart-type>>
```

Syntax rules:

  • The language tag is fixed as GPT-Vis
  • The content area follows the Syntax format rules, and the first line must contain vis <chart-type> (see the full list in Supported Chart Types above)
  • The code block is converted by the Markdown plugin into and rendered on the browser side

Note: In Markdown mode, the content area must include the first line vis <type>, consistent with the plain Syntax mode format.

Code Mode

Generate complete runnable code based on the target framework.

Installation

NPM:

npm install @antv/gpt-vis
import { GPTVis } from '@antv/gpt-vis';

CDN:

<script src="https://unpkg.com/@antv/gpt-vis/dist/umd/index.min.js"></script>

After loading via CDN, access the main class through GPTVis.GPTVis.

Complete HTML Example

<html>
  <head>
    <script src="https://unpkg.com/@antv/gpt-vis/dist/umd/index.min.js"></script>
  </head>
  <body>
    <div id="container"></div>
    <script>
      const gptVis = new GPTVis.GPTVis({
        container: '#container',
        width: 600,
        height: 400,
      });

      gptVis.render(`
vis line
data
  - time 2020
    value 100
  - time 2021
    value 120
title Annual Trend
`);
    </script>
  </body>
</html>

Chart Type Configurations

Common Configuration

All charts include the following fields, which are omitted from the individual chart type definitions below. Each section heading is the type value (e.g. line, column), corresponding to the type column in the chart type table above.

{ type: string; title?: string; theme?: 'default' | 'light' | 'dark' | 'academy'; style?: { backgroundColor?: string; palette?: string[] } }

line / area

{ data: { time: string | number; value: number; group?: string }[]; axisXTitle?: string; axisYTitle?: string; stack?: boolean; style?: { lineWidth?: number } }

stack is only supported by area.

column / bar

{ data: { category: string; value: number; group?: string }[]; axisXTitle?: string; axisYTitle?: string; stack?: boolean; group?: boolean }

pie

Values must not be percentage numbers.

{ data: { category: string; value: number }[]; innerRadius?: number }

Set innerRadius to 0.6 to turn it into a donut chart.

scatter

{ data: { x: number; y: number; group?: string }[]; axisXTitle?: string; axisYTitle?: string }

dual-axes

{ categories: string[]; series: { type: 'line' | 'column'; data: number[]; axisYTitle?: string }[]; axisXTitle?: string; style?: { startAtZero?: boolean } }

histogram

{ data: number[]; binNumber?: number; axisXTitle?: string; axisYTitle?: string }

boxplot / violin

Multiple data points are required for the same category to show distribution.

{ data: { category: string; value: number; group?: string }[]; axisXTitle?: string; axisYTitle?: string; style?: { startAtZero?: boolean } }

radar

{ data: { name: string; value: number; group?: string }[]; align?: boolean }

align: whether to align the scales of all dimensions. Defaults to false (each axis scales independently); when true, all axes share the same maximum value, suitable for comparing absolute values across multiple series.

funnel

{ data: { category: string; value: number; }[]; }

waterfall

Values can be negative to indicate decrease. palette is a color array in the order [positive color, negative color, total color].

{ data: { category: string; value: number }[]; axisXTitle?: string; axisYTitle?: string; style?: { palette?: string[] } }

liquid

percent ranges from 0 to 1.

{ percent: number; shape?: 'rect' | 'circle' | 'pin' | 'triangle' }

word-cloud

{ data: { text: string; value: number; }[]; }

venn

Intersections use commas to separate set identifiers: sets: "A,B". label is used to display the name of the corresponding set on the chart.

{ data: { sets: string | string[]; value: number; label?: string }[] }

treemap

type TreeNode = { name: string; value: number; children?: TreeNode[] };
{ data: TreeNode[] }

sankey

{ data: { source: string; target: string; value: number }[]; nodeAlign?: 'left' | 'center' | 'right' | 'justify' }

flow-diagram / network-graph

source/target reference node name values.

type GraphData = { nodes: { name: string }[]; edges: { source: string; target: string; name?: string }[] };

// flow-diagram
{ data: GraphData }

// network-graph
{ data: GraphData; layout?: 'force' | 'circular' | 'grid' | 'radial' | 'concentric' | 'dagre' }

mindmap / indented-tree / organization-chart

type TreeData = { name: string; children?: TreeData[] };

// mindmap
{ data: TreeData; direction?: 'H' | 'LR' | 'RL' }

// indented-tree
{ data: TreeData; direction?: 'LR' | 'RL' | 'H' }

// organization-chart
type OrganizationChartData = {
  name: string;
  description?: string;
  children?: OrganizationChartData[];
};

{ data: OrganizationChartData }

mindmap defaults to 'H'; indented-tree defaults to 'LR'.

fishbone-diagram

type FishboneNode = { name: string; children?: FishboneNode[] };
{ data: FishboneNode; style?: { texture?: 'rough' | 'default' } }

texture: 'rough' produces a hand-drawn style.

table

{ data: Record<string, string | number>[]; }

summary

summary is completely different from other chart types: it does not use Syntax/JSON configuration, but instead uses T8 syntax (Markdown + semantic annotations).

⚠️ Before generating a summary, you must: first read references/summary.md to get the T8 syntax rules, the complete entity type list, attribute field definitions, generation requirements, and examples, and only then generate the content. Skipping this step will cause syntax errors.

Best Practices

  1. Keep pie charts to no more than 5 categories; merge the rest into "Other" or switch to a bar chart
  2. Do not use pie charts for trends, and do not use line charts for unordered categories
  3. Value fields must be numeric; category fields must be text
  4. Distributions of continuous values (e.g. salaries, scores, ages) must use histogram
  5. Multi-dimensional field mapping: when there are two categorical dimensions, write the x-axis dimension as time/category and the other as group
  6. In Syntax mode, prefer the Syntax format (streaming-friendly)
  7. In Code mode, default to the HTML + CDN approach (zero installation); use the npm approach only when the user specifies a framework

Infographics convert data, information, and knowledge into perceptible visual language. They combine visual design with data visualization, compressing complex information with intuitive symbols to help audiences quickly understand and remember key points.

Infographic = Information Structure + Visual Expression

This task uses AntV Infographic to create visual infographics.

Before starting the task, you need to understand the AntV Infographic syntax specifications, including template list, data structure, themes, etc.

Specifications

AntV Infographic Syntax

AntV Infographic syntax is a custom DSL used to describe infographic rendering configurations. It uses indentation to describe information, has strong robustness, and is convenient for AI streaming output and infographic rendering. It mainly contains the following information:

  1. template: Use templates to express the text information structure.
  2. data: Infographic data, including title, desc, data items, etc. Data items typically contain fields such as label, desc, icon, etc.
  3. theme: Theme contains style configurations such as palette, font, etc.

For example:

infographic list-row-horizontal-icon-arrow
data
  title Title
  desc Description
  lists
    - label Label
      value 12.5
      desc Explanation
      icon document text
theme
  palette #3b82f6 #8b5cf6 #f97316

Syntax Specifications

  • The first line must be infographic <template-name>, template selected from the list below (see "Available Templates" section).

  • Use data / theme blocks, with two-space indentation within blocks.

  • Key-value pairs use "key space value"; arrays use - as entry prefix.

  • icon uses icon keywords (e.g., star fill).

  • data should contain title/desc + template-specific main data field (not necessarily items).

  • Main data field selection (use only one, avoid mixing):

    • list-* → lists
    • sequence-* → sequences (optional order asc|desc)
    • compare-* → compares (supports children for grouped comparisons), can contain multiple comparison items
    • hierarchy-structure → items (each item corresponds to an independent hierarchy, each level can contain sub-items, can be nested up to 3 levels)
    • hierarchy-* → single root (tree structure, nested through children);
    • relation-* → nodes + relations; simple relation diagrams can omit nodes, using arrow syntax in relations
    • chart-* → values (numeric statistics, optional category)
    • Use items as fallback when uncertain
  • compare-binary-* / compare-hierarchy-left-right-* binary templates: must have two root nodes, all comparison items hang under these two root nodes' children

  • hierarchy-*: use single root, nested through children (do not repeat root)

  • theme is used to customize themes (palette, font, etc.) For example: dark theme + custom color scheme

    infographic list-row-horizontal-icon-arrow
    theme dark
      palette
        - #61DDAA
        - #F6BD16
        - #F08BB4
    
  • Use theme.base.text.font-family to specify font, such as handwriting style 851tegakizatsu

  • Use theme.stylize to select built-in styles and pass parameters Common styles:

    • rough: hand-drawn effect
    • pattern: pattern fill
    • linear-gradient / radial-gradient: linear/radial gradient

    For example: hand-drawn style (rough)

    infographic list-row-horizontal-icon-arrow
    theme
      stylize rough
      base
        text
          font-family 851tegakizatsu
    
  • Do not output JSON, Markdown, or explanatory text

Data Syntax Examples

Data syntax examples by template category (use corresponding fields, avoid adding items simultaneously):

  • list-* templates
infographic list-grid-badge-card
data
  title Feature List
  lists
    - label Fast
      icon flash fast
    - label Secure
      icon secure shield check
  • sequence-* templates
infographic sequence-steps-simple
data
  sequences
    - label Step 1
    - label Step 2
    - label Step 3
  order asc
  • hierarchy-* templates
infographic hierarchy-structure
data
  root
    label Company
    children
      - label Dept A
      - label Dept B
  • compare-* templates
infographic compare-swot
data
  compares
    - label Strengths
      children
        - label Strong brand
        - label Loyal users
    - label Weaknesses
      children
        - label High cost
        - label Slow release

Quadrant diagram

infographic compare-quadrant-quarter-simple-card
data
  compares
    - label High Impact & Low Effort
    - label High Impact & High Effort
    - label Low Impact & Low Effort
    - label Low Impact & High Effort
  • chart-* templates
infographic chart-column-simple
data
  values
    - label Visits
      value 1280
    - label Conversion
      value 12.4
  • relation-* templates

Edge label syntax: A -label-> B or A -->|label| B

infographic relation-dagre-flow-tb-simple-circle-node
data
  nodes
    - id A
      label Node A
    - id B
      label Node B
  relations
    A - approves -> B
    A -->|blocks| B
  • Fallback items example
infographic list-row-horizontal-icon-arrow
data
  items
    - label Item A
      desc Description
      icon sun
    - label Item B
      desc Description
      icon moon

Available Templates

  • chart-bar-plain-text
  • chart-column-simple
  • chart-line-plain-text
  • chart-pie-compact-card
  • chart-pie-donut-pill-badge
  • chart-pie-donut-plain-text
  • chart-pie-plain-text
  • chart-wordcloud
  • compare-binary-horizontal-badge-card-arrow
  • compare-binary-horizontal-simple-fold
  • compare-binary-horizontal-underline-text-vs
  • compare-hierarchy-left-right-circle-node-pill-badge
  • compare-quadrant-quarter-circular
  • compare-quadrant-quarter-simple-card
  • compare-swot
  • hierarchy-mindmap-branch-gradient-capsule-item
  • hierarchy-mindmap-level-gradient-compact-card
  • hierarchy-structure
  • hierarchy-tree-curved-line-rounded-rect-node
  • hierarchy-tree-tech-style-badge-card
  • hierarchy-tree-tech-style-capsule-item
  • list-column-done-list
  • list-column-simple-vertical-arrow
  • list-column-vertical-icon-arrow
  • list-grid-badge-card
  • list-grid-candy-card-lite
  • list-grid-ribbon-card
  • list-row-horizontal-icon-arrow
  • list-sector-plain-text
  • list-zigzag-down-compact-card
  • list-zigzag-down-simple
  • list-zigzag-up-compact-card
  • list-zigzag-up-simple
  • relation-dagre-flow-tb-animated-badge-card
  • relation-dagre-flow-tb-animated-simple-circle-node
  • relation-dagre-flow-tb-badge-card
  • relation-dagre-flow-tb-simple-circle-node
  • sequence-ascending-stairs-3d-underline-text
  • sequence-ascending-steps
  • sequence-circular-simple
  • sequence-color-snake-steps-horizontal-icon-line
  • sequence-cylinders-3d-simple
  • sequence-filter-mesh-simple
  • sequence-funnel-simple
  • sequence-horizontal-zigzag-underline-text
  • sequence-mountain-underline-text
  • sequence-pyramid-simple
  • sequence-roadmap-vertical-plain-text
  • sequence-roadmap-vertical-simple
  • sequence-snake-steps-compact-card
  • sequence-snake-steps-simple
  • sequence-snake-steps-underline-text
  • sequence-stairs-front-compact-card
  • sequence-stairs-front-pill-badge
  • sequence-timeline-rounded-rect-node
  • sequence-timeline-simple
  • sequence-zigzag-pucks-3d-simple
  • sequence-zigzag-steps-underline-text

Template Selection Recommendations:

  • Strict sequence (process/steps/development trend) → sequence-*
    • Timeline → sequence-timeline-*
    • Staircase diagram → sequence-stairs-*
    • Roadmap → sequence-roadmap-vertical-*
    • Zigzag path → sequence-zigzag-*
    • Circular progress → sequence-circular-simple
    • Colorful snake steps → sequence-color-snake-steps-*
    • Pyramid → sequence-pyramid-simple
  • Opinion listing → list-row-* or list-column-*
  • Binary comparison (pros/cons) → compare-binary-*
  • SWOT → compare-swot
  • Hierarchical structure (tree diagram) → hierarchy-tree-*
  • Data charts → chart-*
  • Quadrant analysis → compare-quadrant-*
  • Grid list (key points) → list-grid-*
  • Relationship display → relation-*
  • Word cloud → chart-wordcloud
  • Mind map → hierarchy-mindmap-*

Example

Creating an Internet technology evolution infographic

infographic list-row-horizontal-icon-arrow
data
  title Internet Technology Evolution
  desc From Web 1.0 to AI era, key milestones
  lists
    - time 1991
      label Web 1.0
      desc Tim Berners-Lee published the first website, opening the Internet era
      icon web
    - time 2004
      label Web 2.0
      desc Social media and user-generated content become mainstream
      icon account multiple
    - time 2007
      label Mobile
      desc iPhone released, smartphone changes the world
      icon cellphone
    - time 2015
      label Cloud Native
      desc Containerization and microservices architecture are widely used
      icon cloud
    - time 2020
      label Low Code
      desc Visual development lowers the technology threshold
      icon application brackets
    - time 2023
      label AI Large Model
      desc ChatGPT ignites the generative AI revolution
      icon brain

Generation Process

Step 1: Understand User Requirements

Before creating an infographic, first understand the user's needs and the information they want to express, in order to determine the template and data structure.

If the user provides a clear content description, it should be broken down into a clear and concise structure.

Otherwise, clarification from the user is needed (e.g., "Please provide a clear and concise content description.", "Which template do you want to use?")

  • Extract key information structure (title, desc, items, etc.).
  • Clarify required data fields (title, desc, items, label, value, icon, etc.).
  • Select appropriate template.
  • Describe infographic content using AntV Infographic syntax {syntax}.

Key Note: Must respect the language of user input. For example, if the user inputs in Chinese, the text in the syntax must also be in Chinese.

Step 2: Render the Infographic

When you have the final AntV Infographic syntax, you can generate a complete HTML file following these steps:

  1. Create a complete HTML file with the following structure:
    • DOCTYPE and HTML meta (charset: utf-8)
    • Title: {title} - Infographic
    • Include AntV Infographic script: https://unpkg.com/@antv/infographic@latest/dist/infographic.min.js
    • Create container div with id container
    • Initialize Infographic (width: '100%', height: '100%')
    • Replace {title} with actual title
    • Replace {syntax} with actual AntV Infographic syntax
    • Add SVG export functionality: const svgDataUrl = await infographic.toDataURL({ type: 'svg' });

Reference HTML template:

<div id="container"></div>
<script src="https://unpkg.com/@antv/infographic@latest/dist/infographic.min.js"></script>
<script>
 const infographic = new AntVInfographic.Infographic({
    container: '#container',
    width: '100%',
    height: '100%',
  });
  document.fonts?.ready.then(() => {
    infographic.render(`{syntax}`);
  }).catch((error) => {
    console.error('Error waiting for fonts to load:', error);
    infographic.render(`{syntax}`);
  });
</script>
  1. Use the Write tool to generate HTML file, named as <title>-infographic.html

  2. Show to user:

    • Generate file path and prompt: "Open directly with a browser to view and save as SVG"
    • Output syntax and prompt: "Tell me if you need to adjust template/colors/content"

Note: The HTML file must include:

  • SVG export via export button
  • Container is responsive, both width and height are 100%

Narrative Text Visualization Skill

This skill provides a workflow for transforming data into structured narrative text visualizations using T8 Syntax - a declarative Markdown-like language for creating data narratives with semantic entity annotations.

What is T8

T8 is a text visualization solution under the AntV technology stack designed specifically for insight-based narrative text display. Instead of manually constructing DOM elements, you write simple, human-readable syntax that describes your data narrative.

Key Features:

  • LLM-Friendly: The syntax is intuitive and can be easily generated by AI models
  • Declarative & Readable: Write what you want, not how to build it
  • Framework Agnostic: Works with React, Vue, or vanilla JavaScript
  • Standardized Styling: Professional appearance by default
  • Built-in Data Visualizations: Mini charts (pie, line) are native to the syntax
  • Lightweight: Less than 20KB before gzip

Workflow

To generate narrative text visualizations, follow these steps:

1. Understand the Requirements

Analyze the user's request to determine:

  • The topic or data to be analyzed
  • The type of narrative needed (report, summary, article)
  • The key insights to highlight
  • Any specific data sources or metrics

2. Generate T8 Syntax Content

Create narrative text using T8 Syntax following the specification below. The content must include:

  • Proper document structure (headings, paragraphs, lists)
  • Entity annotations for all meaningful data points
  • Appropriate metadata for entities (origin, assessment, etc.)

3. Generate Frontend Code

Create HTML, React, or Vue code to render the T8 content based on user's preferred framework.

4. Validate Output

Ensure:

  • All data is from authentic sources
  • Minimum content length (800 words or equivalent)
  • Proper entity annotations throughout
  • Clear structure and logical flow

T8 Syntax Specification

T8 Syntax is a Markdown-like language for creating narrative text with semantic entity annotations. It makes data analysis reports more expressive and visually appealing.

Document Structure

Headings (6 levels)

Use standard Markdown heading syntax:

# Level 1 Heading (Main Title)
## Level 2 Heading (Section)
### Level 3 Heading (Subsection)
#### Level 4 Heading
##### Level 5 Heading
###### Level 6 Heading

Rules:

  • Each heading must be on its own line
  • Add one space after the # symbols
  • Headings create visual hierarchy in the rendered output
Paragraphs

Regular text paragraphs are separated by blank lines:

This is the first paragraph with some content.

This is the second paragraph, separated by a blank line.

Rules:

  • Paragraphs can span multiple lines
  • Use blank lines to separate distinct paragraphs
  • Text within a paragraph flows naturally
Lists

T8 Syntax supports both unordered and ordered lists.

Unordered Lists:

- First item
- Second item
- Third item

Ordered Lists:

1. First step
2. Second step
3. Third step

Rules:

  • Each list item must be on its own line
  • Add one space after the bullet marker (-, *) or number
  • Lists can contain entities and text formatting

Text Formatting

T8 Syntax supports inline text formatting using Markdown syntax:

Bold Text: This is **bold text** that stands out.

Italic Text: This is *italic text* for emphasis.

Underline Text: This is __underlined text__ for importance.

Links: Visit [our website](https://example.com) for more information.

Rules:

  • Formatting markers must be balanced (opening and closing)
  • Formatting can be combined with entities
  • Links use [text](URL) syntax where URL starts with http://, https://, or /

Entity Annotation Syntax

The core feature of T8 Syntax is entity annotation - marking specific data points with semantic meaning and metadata.

Basic Entity Syntax
[displayText](entityType)
  • displayText: The text shown to readers
  • entityType: The semantic type of this entity

Example:

The [sales revenue](metric_name) reached [¥1.5 million](metric_value) this quarter.
Entity with Metadata
[displayText](entityType, key1=value1, key2=value2, key3="string value")

Metadata Rules:

  • Separate multiple metadata fields with commas
  • Numbers and booleans: write directly (e.g., origin=1500000, active=true)
  • Strings: wrap in double quotes (e.g., unit="元", region="Asia")

Example:

Revenue grew by [15.3%](ratio_value, origin=0.153, assessment="positive") compared to last year.

Entity Types Reference

Use these entity types to annotate different kinds of data:

Entity TypeDescriptionWhen to UseExamples
metric_nameName of a metric or KPIWhen mentioning what you're measuring"revenue", "user count", "market share"
metric_valuePrimary metric valueThe main number/value being reported"¥1.5 million", "50,000 users", "250 units"
other_metric_valueSecondary or supporting metric valueAdditional metrics that provide context"average order value: $120"
delta_valueAbsolute change/differenceWhen showing numeric change between periods"+1,200 units", "-$50K", "increased by 500"
ratio_valuePercentage change/rateWhen showing percentage change"+15.3%", "-5.2%", "grew 23%"
contribute_ratioContribution percentageWhen showing what % something contributes"accounts for 45%", "represents 30% of total"
trend_descTrend descriptionDescribing direction/pattern of change"steadily rising", "declining trend", "stable"
dim_valueDimensional value/categoryGeographic, categorical, or segmentation data"North America", "Enterprise segment", "Q3"
time_descTime period or timestampWhen specifying when something occurred"Q3 2024", "January-March", "fiscal year 2023"
proportionProportion or ratioWhen expressing parts of a whole"3 out of 5", "60% of customers"
rankRanking or positionWhen indicating order or position in a list"ranked 1st", "top 3", "5th place"
differenceComparative differenceWhen highlighting difference between two items"difference of $50K", "gap of 200 units"
anomalyUnusual or unexpected valueWhen pointing out outliers or anomalies"unusual spike", "unexpected drop"
associationRelationship or correlationWhen describing connections between metrics"strongly correlated", "linked to", "related"
distributionData distribution patternWhen describing how data is spread"evenly distributed", "concentrated in", "spread across"
seasonalitySeasonal pattern or trendWhen describing recurring seasonal patterns"seasonal peak", "holiday period", "Q4 surge"

Common Metadata Fields

Add these optional fields to provide richer data context:

origin (number)

The raw numerical value behind the displayed text.

Examples:

  • [¥1.5M](metric_value, origin=1500000)
  • [23.7%](ratio_value, origin=0.237)
  • [5.2K users](metric_value, origin=5200)
  • [3 out of 4](proportion, origin=0.75)

Why use it: Enables data visualization, sorting, and calculations

assessment (string)

Evaluates whether a change is positive, negative, or neutral.

Valid values: "positive", "negative", "equal", "neutral"

Examples:

  • [increased 15%](ratio_value, assessment="positive")
  • [dropped 8%](ratio_value, assessment="negative")
  • [remained flat](trend_desc, assessment="equal")

Why use it: Enables visual indicators (colors, icons) for good/bad trends

unit (string)

The unit of measurement for the value.

Examples:

  • [¥1,500,000](metric_value, unit="元", origin=1500000)
  • [150](metric_value, unit="units")
detail (any)

Additional context or breakdown data for chart rendering. Required for certain entity types.

Required for these entity types:

  • rank: Array of numbers representing ranking data
    • Example: [top performer](rank, detail=[5, 8, 12, 15, 20])
  • difference: Array of numbers showing comparative values
    • Example: [gap narrowing](difference, detail=[100, 80, 60, 40])
  • anomaly: Array of numbers highlighting outliers
    • Example: [unusual spike](anomaly, detail=[10, 12, 11, 45, 13])
  • association: Array of {x, y} objects for correlation data
    • Example: [strong correlation](association, detail=[{"x":1,"y":2},{"x":2,"y":4},{"x":3,"y":6}])
  • distribution: Array of numbers showing data spread
    • Example: [uneven distribution](distribution, detail=[5, 15, 45, 25, 10])
  • seasonality: Object with data array and optional range
    • Example: [Q4 peak](seasonality, detail={"data":[10,12,15,30],"range":[0,40]})

Optional for other types:

  • [steady growth](trend_desc, detail=[100, 120, 145, 180, 210])

Data Requirements

Critical: All data must be from publicly authentic sources:

  • Official announcements/financial reports
  • Authoritative media (Reuters, Bloomberg, TechCrunch, etc.)
  • Industry research institutions (IDC, Canalys, Counterpoint Research, etc.)
  • Never use fictional, AI-guessed, or simulated data
  • Use specific numbers (e.g., "146 million units", "7058 units"), not vague approximations

Complete T8 Syntax Example

# 2024 Smartphone Market Analysis

## Market Overview

Global [smartphone shipments](metric_name) reached [1.2 billion units](metric_value, origin=1200000000) in [2024](time_desc), showing a [modest decline of 2.1%](ratio_value, origin=-0.021, assessment="negative") year-over-year.

The **premium segment** (devices over $800) showed *remarkable* [resilience](trend_desc, assessment="positive"), growing by [5.8%](ratio_value, origin=0.058, assessment="positive"). [Average selling price](other_metric_value) was [$420](metric_value, origin=420, unit="USD").

## Key Findings

1. [Asia-Pacific](dim_value) remains the __largest market__
2. [Premium devices](dim_value) showed **strong growth**
3. Budget segment faced *headwinds*

## Regional Breakdown

### Asia-Pacific

[Asia-Pacific](dim_value) remains the largest market with [680 million units](metric_value, origin=680000000) shipped, though this represents a [decline of 180 million units](delta_value, origin=-180000000, assessment="negative") from the previous year.

Key markets:
- [China](dim_value): [320M units](metric_value, origin=320000000) - down [8.5%](ratio_value, origin=-0.085, assessment="negative"), [ranked 1st](rank, detail=[320, 180, 90, 65, 45]) globally, accounting for [47%](contribute_ratio, origin=0.47, assessment="positive") of regional sales
- [India](dim_value): [180M units](metric_value, origin=180000000) - up [12.3%](ratio_value, origin=0.123, assessment="positive"), [ranked 2nd](rank, detail=[320, 180, 90, 65, 45])
- [Southeast Asia](dim_value): [180M units](metric_value, origin=180000000) - [stable](trend_desc, assessment="equal")

For detailed methodology, visit [our research page](https://example.com/methodology).

Using T8 in HTML, React, and Vue

Using in HTML (via CDN)

<!DOCTYPE html>
<html lang="en">
<head>
  <meta charset="UTF-8">
  <meta name="viewport" content="width=device-width, initial-scale=1.0">
  <title>T8 Narrative Text</title>
</head>
<body>
  <div id="container"></div>

  <!-- Import T8 from unpkg CDN -->
  <script src="https://unpkg.com/@antv/t8/dist/t8.min.js"></script>
  
  <script>
    // T8 is available as a global variable
    const { Text } = window.T8;

    // Initialize T8 instance
    const text = new Text(document.getElementById('container'));

    // Render narrative text using T8 Syntax
    const narrativeText = `
# Sales Report

This quarter, [bookings](metric_name) are higher than usual. They are [¥348k](metric_value, origin=348.12).

[Bookings](metric_name) are up [¥180.3k](delta_value, assessment="positive") relative to the same time last quarter.
    `;

    text.theme('light').render(narrativeText);
  </script>
</body>
</html>

Installation:

npm install @antv/t8
# or
yarn add @antv/t8

Using in React

import { Text } from '@antv/t8';
import { useEffect, useRef } from 'react';

function T8Component() {
  const containerRef = useRef<HTMLDivElement>(null);

  useEffect(() => {
    if (!containerRef.current) return;

    // Initialize T8 instance
    const text = new Text(containerRef.current);

    // Render narrative text using T8 Syntax
    const narrativeText = `
# Sales Report

This quarter, [bookings](metric_name) are higher than usual. They are [¥348k](metric_value, origin=348.12).

[Bookings](metric_name) are up [¥180.3k](delta_value, assessment="positive") relative to the same time last quarter.
    `;

    text.theme('light').render(narrativeText);

    // Cleanup on unmount
    return () => {
      text.unmount();
    };
  }, []);

  return <div ref={containerRef} />;
}

export default T8Component;

Using in Vue 3

<template>
  <div ref="containerRef"></div>
</template>

<script setup lang="ts">
import { Text } from '@antv/t8';
import { ref, onMounted, onBeforeUnmount } from 'vue';

const containerRef = ref<HTMLDivElement>();
let textInstance: Text | null = null;

onMounted(() => {
  if (!containerRef.value) return;

  // Initialize T8 instance
  textInstance = new Text(containerRef.value);

  // Render narrative text using T8 Syntax
  const narrativeText = `
# Sales Report

This quarter, [bookings](metric_name) are higher than usual. They are [¥348k](metric_value, origin=348.12).

[Bookings](metric_name) are up [¥180.3k](delta_value, assessment="positive") relative to the same time last quarter.
  `;

  textInstance.theme('light').render(narrativeText);
});

onBeforeUnmount(() => {
  if (textInstance) {
    textInstance.unmount();
  }
});
</script>

Using in Vue 2

<template>
  <div ref="container"></div>
</template>

<script>
import { Text } from '@antv/t8';

export default {
  name: 'T8Component',
  data() {
    return {
      textInstance: null,
    };
  },
  mounted() {
    // Initialize T8 instance
    this.textInstance = new Text(this.$refs.container);

    // Render narrative text using T8 Syntax
    const narrativeText = `
# Sales Report

This quarter, [bookings](metric_name) are higher than usual. They are [¥348k](metric_value, origin=348.12).

[Bookings](metric_name) are up [¥180.3k](delta_value, assessment="positive") relative to the same time last quarter.
    `;

    this.textInstance.theme('light').render(narrativeText);
  },
  beforeDestroy() {
    if (this.textInstance) {
      this.textInstance.unmount();
    }
  },
};
</script>

Writing Guidelines and Best Practices

Content Requirements

  1. Minimum Length: No less than 800 words (adjust based on data complexity)
  2. Structure: Clear hierarchy with logical flow between sections
  3. Analysis: Don't just list numbers - explain their significance and context
  4. Tone: Natural, fluent, objective, and professional
  5. Entity Usage: Annotate ALL meaningful data points - metrics, values, trends, times, changes, percentages

Entity Annotation Best Practices

  1. Be Comprehensive: Mark all quantitative data, not just major figures
  2. Use Appropriate Types: Choose the entity type that best describes the semantic meaning
  3. Add Metadata: Include origin, assessment, and other relevant fields when applicable
  4. Natural Flow: Entities should blend seamlessly into readable prose

What to Annotate

✅ DO annotate:

  • All numeric values (revenue, counts, measurements)
  • All percentages (changes, contributions, proportions)
  • Metric names and KPIs
  • Time periods
  • Geographic regions and categories
  • Trend descriptions
  • Comparisons and changes

❌ DON'T annotate:

  • Generic text without specific data meaning
  • Connecting phrases and transitions
  • Context that doesn't represent measurable concepts

Output Format

When generating T8 Syntax content for the user:

  1. Output the T8 Syntax content directly without wrapping in code blocks
  2. Provide the frontend code (HTML/React/Vue) based on user preference
  3. Ensure all entities are properly annotated with appropriate metadata
  4. Verify that content meets minimum length and quality requirements

The rendered output provides:

  • Rich semantic markup for data entities
  • Interactive entity highlighting
  • Clear visual hierarchy
  • Professional report-style formatting
  • Responsive design for all devices

Reference Links

X6 v3 Graph Editor

Overview

X6 v3 is AntV's diagram editing engine for flowcharts, DAGs, ER diagrams, org charts, and other interactive node-edge editors. Unlike G2/G6, X6 uses an imperative API — you create a Graph instance, then call graph.addNode(), graph.addEdge(), and register plugins via graph.use().

import { Graph } from '@antv/x6';

const graph = new Graph({
  container: 'container',
  background: { color: '#F2F7FA' },
});

const source = graph.addNode({
  shape: 'rect',
  x: 40, y: 40, width: 100, height: 40,
  label: 'Source',
  attrs: { body: { stroke: '#8f8f8f', strokeWidth: 1, fill: '#fff', rx: 6, ry: 6 } },
});

const target = graph.addNode({
  shape: 'rect',
  x: 300, y: 200, width: 100, height: 40,
  label: 'Target',
  attrs: { body: { stroke: '#8f8f8f', strokeWidth: 1, fill: '#fff', rx: 6, ry: 6 } },
});

graph.addEdge({ source, target, attrs: { line: { stroke: '#8f8f8f', strokeWidth: 1 } } });
graph.centerContent();

CDN Usage

<script src="https://unpkg.com/@antv/x6@3/dist/x6.js"></script>
<script>
  const graph = new X6.Graph({
    container: 'container',
    background: { color: '#F2F7FA' },
  });
  const source = graph.addNode({
    shape: 'rect',
    x: 40, y: 40, width: 100, height: 40,
    label: 'Source',
    attrs: { body: { stroke: '#8f8f8f', strokeWidth: 1, fill: '#fff', rx: 6, ry: 6 } },
  });
  const target = graph.addNode({
    shape: 'rect',
    x: 300, y: 200, width: 100, height: 40,
    label: 'Target',
    attrs: { body: { stroke: '#8f8f8f', strokeWidth: 1, fill: '#fff', rx: 6, ry: 6 } },
  });
  graph.addEdge({ source, target, attrs: { line: { stroke: '#8f8f8f', strokeWidth: 1 } } });
  graph.centerContent();
</script>

Content Retrieval Service

When using AntV X6 for data visualization, if you need to understand the concepts, usage, API, examples, and other aspects of X6 v3, you can use the provided context retrieval service. When using the skill, content is retrieved via an antv HTTP API server using GET requests.

  • Host: https://sive.antv.antgroup.com
  • Endpoint: /api/v1/context/retrieve
  • Method: GET
  • Parameters: query, library, topK, content, maxTokens

Retrieve skills by query (hybrid search = FTS + vector + RRF fusion). Constraints docs are indexed as regular skill documents and will appear in search results naturally.

ParameterTypeRequiredDescription
querystring✅Search keywords, e.g. flowchart stencil port
librarystring✅Library name: g2, g6, x6
topKnumberNumber of results to return (default: 5)
contentbooleanReturn full reference doc markdown (default: true)
maxTokensnumberMax tokens per result (default: unlimited)
curl "https://sive.antv.antgroup.com/api/v1/context/retrieve?query=flowchart+stencil+port&library=x6"

Critical Rules

MUST: graph.render() does NOT exist in X6 v3

// ❌ WRONG — graph.render() is G6 API, not X6
const graph = new Graph({ container: 'container' });
graph.render();

// ✅ CORRECT — X6 auto-renders on addNode/addEdge/fromJSON
const graph = new Graph({ container: 'container', background: { color: '#F2F7FA' } });
graph.addNode({ shape: 'rect', x: 40, y: 40, width: 100, height: 40 });

MUST: Use string literal container: 'container' — no variable declaration

// ❌ WRONG — declaring container variable is forbidden
const container = document.getElementById('container');
const graph = new Graph({ container });

// ✅ CORRECT — string literal, runtime auto-resolves
const graph = new Graph({ container: 'container', background: { color: '#F2F7FA' } });

MUST: Register plugins before using their methods

// ❌ WRONG — calling plugin method without registration
graph.toPNG();       // Error: method not found
graph.select();      // Error: method not found

// ✅ CORRECT — register first, then call
import { Graph, Export, Selection } from '@antv/x6';
const graph = new Graph({ container: 'container', background: { color: '#F2F7FA' } });
graph.use(new Export());
graph.use(new Selection({ enabled: true, rubberband: true }));
// Now graph.toPNG() and graph.select() are available

MUST: Only 11 plugin classes exist — NOT constructor options

✅ Plugin class (import + graph.use)❌ NOT a plugin (constructor option)
Clipboard, Dnd, Export, History, Keyboard, MiniMap, Scroller, Selection, Snapline, Stencil, Transformmousewheel, embedding, panning, connecting, translating, interacting, background, grid
// ❌ WRONG — importing constructor option as "plugin"
import { Graph, Embedding } from '@antv/x6';  // Embedding doesn't exist!
graph.use(new Embedding());                   // Error: not a constructor

// ✅ CORRECT — embedding is a Graph constructor option
import { Graph, Selection } from '@antv/x6';
const graph = new Graph({
  container: 'container',
  embedding: { enabled: true, findParent: 'bbox' },
  mousewheel: { enabled: true, zoomAtMousePosition: true, modifiers: ['ctrl'] },
});
graph.use(new Selection({ enabled: true, rubberband: true }));

MUST: All used classes MUST appear in import statement

// ❌ WRONG — Selection used but not imported
import { Graph } from '@antv/x6';
graph.use(new Selection({...}));  // falls back to window.Selection → Illegal constructor

// ✅ CORRECT — every used class imported
import { Graph, Selection, Keyboard, History } from '@antv/x6';
graph.use(new Selection({ enabled: true, rubberband: true }));
graph.use(new Keyboard({ enabled: true }));
graph.use(new History({ enabled: true }));

MUST: Always call graph.centerContent() after adding nodes/edges

// ❌ WRONG — no centerContent, content drifts to top-left
graph.addNode({ ... });
graph.addEdge({ ... });

// ✅ CORRECT — content centered after all additions
graph.addNode({ ... });
graph.addEdge({ ... });
graph.centerContent();
// OR: graph.zoomToFit({ padding: 20, maxScale: 1 }) — but NOT both

MUST: Always set background color, default node/edge style

// ❌ WRONG — no background, no default styles
const graph = new Graph({ container: 'container' });

// ✅ CORRECT — mandatory background + default styles
const graph = new Graph({ container: 'container', background: { color: '#F2F7FA' } });
graph.addNode({
  shape: 'rect', x: 40, y: 40, width: 100, height: 40,
  label: 'Node',
  attrs: { body: { stroke: '#8f8f8f', strokeWidth: 1, fill: '#fff', rx: 6, ry: 6 } },
});
graph.addEdge({
  source: 'node-1', target: 'node-2',
  attrs: { line: { stroke: '#8f8f8f', strokeWidth: 1 } },
});

MUST: mousewheel, panning, Selection.rubberband — use modifiers to avoid conflicts

// ❌ WRONG — panning and mousewheel both grab scroll events
const graph = new Graph({
  panning: { enabled: true },
  mousewheel: { enabled: true },
});
graph.use(new Selection({ enabled: true, rubberband: true }));

// ✅ CORRECT — modifiers separate the interactions
const graph = new Graph({
  panning: { enabled: true, eventTypes: ['leftMouseDown'], modifiers: 'shift' },
  mousewheel: { enabled: true, zoomAtMousePosition: true, modifiers: ['ctrl'] },
});
graph.use(new Selection({ enabled: true, rubberband: true }));

MUST: Output pure JavaScript — NO TypeScript syntax

// ❌ WRONG — TypeScript syntax
private width: number = 100;
const node: Node = graph.addNode({...}) as Node;

// ✅ CORRECT — pure JavaScript only
const node = graph.addNode({ shape: 'rect', x: 40, y: 40 });

MUST: Shape.HTML.register for HTML nodes — NOT class extends Node

// ❌ WRONG — class-based HTML node (2.x pattern)
class MyNode extends Node { ... }

// ✅ CORRECT — Shape.HTML.register (3.x pattern)
import { Graph, Shape } from '@antv/x6';
Shape.HTML.register({
  shape: 'my-html',
  effect: ['data'],
  html(node) {
    const div = document.createElement('div');
    div.innerHTML = node.getData().content || '';
    return div;
  },
});

Quick Reference

User IntentRetrieve Query
Graph init, container, backgroundGET /api/v1/context/retrieve?query=graph+init+container+background&library=x6
Flowchart / approval flowGET /api/v1/context/retrieve?query=flowchart+approval&library=x6
DAG / data pipelineGET /api/v1/context/retrieve?query=DAG+pipeline+port&library=x6
ER diagram / entity relationshipGET /api/v1/context/retrieve?query=ER+diagram+entity+relationship&library=x6
Lineage / data lineage graphGET /api/v1/context/retrieve?query=lineage+data+lineage&library=x6
Org chart / hierarchyGET /api/v1/context/retrieve?query=org+chart+hierarchy&library=x6
UML class diagramGET /api/v1/context/retrieve?query=UML+class+diagram&library=x6
Node config / custom nodeGET /api/v1/context/retrieve?query=node+custom+shape+rect+circle&library=x6
Edge config / router / connectorGET /api/v1/context/retrieve?query=edge+router+connector+orth+smooth&library=x6
Ports / connection桩GET /api/v1/context/retrieve?query=ports+connection+layout&library=x6
HTML shape nodeGET /api/v1/context/retrieve?query=html+shape+register&library=x6
Stencil / drag-and-drop panelGET /api/v1/context/retrieve?query=stencil+drag+drop+panel&library=x6
Plugin: Selection, History, ClipboardGET /api/v1/context/retrieve?query=Selection+History+Clipboard+plugin&library=x6
Plugin: MiniMap, Scroller, SnaplineGET /api/v1/context/retrieve?query=MiniMap+Scroller+Snapline+plugin&library=x6
Plugin: Keyboard, Export, TransformGET /api/v1/context/retrieve?query=Keyboard+Export+Transform+plugin&library=x6
Panning / mousewheel / embeddingGET /api/v1/context/retrieve?query=panning+mousewheel+embedding&library=x6
Tools (button-remove, etc.)GET /api/v1/context/retrieve?query=tools+button-remove+hover&library=x6
Events (click,mouseenter,moved)GET /api/v1/context/retrieve?query=events+node+click+mouse&library=x6
Serialization (toJSON, fromJSON)GET /api/v1/context/retrieve?query=serialization+toJSON+fromJSON&library=x6
Animation / gradientGET /api/v1/context/retrieve?query=animation+gradient+defs+marker&library=x6
Group / nesting / embeddingGET /api/v1/context/retrieve?query=group+nesting+embedding+parent+child&library=x6

Dependencies

  • @antv/x6 — X6 v3 diagram editing engine (exports Graph + 11 plugin classes)

请根据用户输入的内容,将数据可视化为图表。

步骤

  1. 分析用户数据和需求,选择最合适的图表类型
  2. 构造符合规范的 JSON 请求体
  3. 使用 curl 工具调用 API 生成图表图片
  4. 将返回的图片 URL 以 Markdown 图片格式输出

图表选择指南

根据用户的数据特征和需求,选择最合适的图表类型:

  • 时间序列:用 line(趋势)或 area(累计趋势);两个不同量纲用 dual-axes
  • 比较类:用 bar(横向分类对比)或 column(纵向分类对比);频率分布用 histogram
  • 占比类:用 pie(比例构成)或 treemap(层级占比)
  • 关系与流程:用 scatter(相关性)、sankey(流向)或 venn(集合重叠)
  • 层级与树形:用 organization-chart 或 mind-map
  • 专用类型:
    • radar:多维度对比
    • funnel:流程阶段转化
    • liquid:百分比/进度
    • word-cloud:文本词频
    • boxplot / violin:统计分布
    • network-graph:复杂节点关系
    • fishbone-diagram:因果分析
    • flow-diagram:流程图
    • spreadsheet:结构化数据表或透视表

API 接口

POST https://antv-studio.alipay.com/api/gpt-vis

请求体为 JSON,必须包含 type 和 source: "chart-visualization-skills" 字段。

示例:

curl -X POST https://antv-studio.alipay.com/api/gpt-vis \
  -H "Content-Type: application/json" \
  -d '{"type":"line","source":"chart-visualization-skills","data":[{"time":"2025-01","value":100}],"title":"示例图表"}'

返回示例:

{"success":true,"resultObj":"https://..."}

将 resultObj 中的 URL 以 Markdown 图片格式输出:![图表](URL)

支持的图表类型

分类图表类型
比较类条形图(bar)、柱状图(column)、瀑布图(waterfall)、双轴图(dual-axes)
趋势类面积图(area)、折线图(line)、散点图(scatter)
分布类箱线图(boxplot)、直方图(histogram)、小提琴图(violin)、漏斗图(funnel)
占比类饼图(pie)、水波图(liquid)、词云(word-cloud)
层级类组织架构图(organization-chart)、思维导图(mind-map)、矩形树图(treemap)、桑基图(sankey)
关系类关系图(network-graph)、韦恩图(venn)
流程类流程图(flow-diagram)、鱼骨图(fishbone-diagram)
多维类雷达图(radar)
表格类表格/透视表(spreadsheet)

通用可选参数

参数类型默认值说明
themestring"default"主题:"default" / "academy" / "dark"
widthnumber600图表宽度
heightnumber400图表高度
titlestring""图表标题
style.texturestring"default"纹理:"default" / "rough"(手绘风格)

带坐标轴的图表还支持:axisXTitle、axisYTitle。

各图表 data 格式

  • area / line: {time: string, value: number, group?: string}[],可选 stack: boolean
  • bar: {category: string, value: number, group?: string}[],可选 group / stack (默认 stack: true)
  • column: {category: string, value: number, group?: string}[],可选 group (默认 true) / stack
  • scatter: {x: number, y: number, group?: string}[]
  • pie: {category: string, value: number}[],可选 innerRadius: number (0-1)
  • radar: {name: string, value: number, group?: string}[]
  • funnel: {category: string, value: number}[]
  • waterfall: {category: string, value?: number, isTotal?: boolean, isIntermediateTotal?: boolean}[]
  • dual-axes: categories: string[], series: {type: "column"|"line", data: number[], axisYTitle?: string}[]
  • histogram: number[],可选 binNumber: number
  • boxplot / violin: {category: string, value: number, group?: string}[]
  • liquid: percent: number (0-1),可选 shape: "circle"|"rect"|"pin"|"triangle"
  • word-cloud: {text: string, value: number}[]
  • sankey: {source: string, target: string, value: number}[],可选 nodeAlign
  • treemap: {name: string, value: number, children?: ...}[] (最深 3 层)
  • venn: {sets: string[], value: number, label?: string}[]
  • network-graph / flow-diagram: {nodes: {name: string}[], edges: {source: string, target: string, name?: string}[]}
  • fishbone-diagram / mind-map: {name: string, children?: ...} (最深 3 层)
  • organization-chart: {name: string, description?: string, children?: ...} (最深 3 层),可选 orient: "horizontal"|"vertical"
  • spreadsheet: Record<string, string | number>[],可选 rows / columns / values(透视表字段)

GPT-Vis 图表可视化技能

GPT-Vis 是一个 AI 原生的可视化库,专为 LLM 时代设计。它采用框架无关的架构,支持 26 种图表类型,通过简单自然的语法让 LLM 能够轻松生成高质量的可视化图表。

步骤

  1. 意图识别:根据用户意图和数据特征选择图表类型
  2. 确定输出模式:根据上下文选择语法模式或代码模式
  3. 生成输出:按所选模式生成内容

支持的图表类型

type 值适用场景
line时间序列趋势
area时间序列趋势+总量
column分类数据对比
bar分类对比(标签长)
pie部分占整体比例
scatter两变量关系
dual-axes不同量级数据对比
histogram连续数值频次分布
boxplot数据分布与异常值
violin数据分布密度
radar多维度对比
funnel流程转化率
waterfall累计增减变化
liquid百分比/进度
word-cloud词频展示
venn集合交并关系
treemap层级数据占比
sankey流量流向
flow-diagram流程步骤
mindmap层级知识梳理
indented-tree树节点层级/目录
network-graph实体间关联关系
organization-chart组织层级
fishbone-diagram根因分析
table表格数据展示
summary内容总结

输出模式

模式一:语法模式(Syntax / JSON)

用于 LLM 应用集成场景,生成图表配置供 GPTVis.render() 消费。支持两种格式:

  • Syntax 格式:类 Markdown 缩进语法,适合流式输出(LLM 逐 token 生成时可实时渲染)
  • JSON 格式:标准 JSON 对象,适合结构化 API 调用

两种格式等价,GPTVis.render() 均可直接接受。

模式二:代码模式

用于用户需要可直接运行的完整代码场景。生成包含安装说明和完整代码的输出。

GPTVis API

GPTVis 是库的统一入口类,负责创建、渲染和销毁图表。

构造函数

new GPTVis(options: VisualizationOptions)

VisualizationOptions:

参数类型必填默认值说明
containerstring | HTMLElement是—CSS 选择器或 DOM 元素
widthnumber否—图表宽度(px)
heightnumber否—图表高度(px)
theme'default' | 'light' | 'dark' | 'academy'否'light'主题
wrapperboolean否false是否显示外层 UI 容器(含标签页、下载、复制等)
localestring否'zh-CN'wrapper 内文案语言

方法

render(config: string | object): void

渲染图表。接受两种输入:

  • Syntax 字符串:以 vis [type] 开头的文本,自动解析为配置对象
  • JSON 配置对象:包含 type 字段的对象
  • 纯文本:不以 vis 开头的字符串会被当作 summary 类型渲染

多次调用 render() 会自动销毁前一个图表再渲染新图表。

destroy(): void

销毁当前图表实例,释放资源。

语法模式:JSON 格式

直接输出符合图表 TypeScript 类型的 JSON 对象,GPTVis.render() 可直接消费。

JSON 示例

{
  "type": "column",
  "data": [
    { "category": "A产品", "value": 30, "group": "线上" },
    { "category": "B产品", "value": 50, "group": "线上" }
  ],
  "title": "产品销量对比",
  "axisXTitle": "产品",
  "axisYTitle": "销量(万)",
  "stack": true,
  "theme": "academy",
  "style": {
    "palette": ["#5B8FF9", "#61DDAA"]
  }
}

语法模式:Syntax 格式

类 Markdown 缩进语法,支持流式渲染。第一行必须是 vis [type]。

语法规则

基本属性 — key value,每行一个:

title 年度趋势
theme dark

对象数组 — data 下每项用 - 开头,子字段缩进:

{ data: { time: string; value: number; }[]; }

对应:

data
  - time 2020
    value 100
  - time 2021
    value 120

纯值数组 — 每项用 - 开头:

{ data: number[] }

对应:

data
  - 10
  - 20

含空格的字符串值 — 用引号(单引号或双引号)包裹;不含空格时可省略引号:

categories
  - "North America"
  - '东南 亚'
  - 欧洲

嵌套对象 — 对象名占一行,子属性缩进:

{ style?: { backgroundColor?: string; palette?: string[] } }

对应:

style
  backgroundColor #f0f2f5
  palette
    - #5B8FF9
    - #61DDAA

递归树形 — children 数组用 - 缩进:

type TreeData = { name: string; children?: TreeData[] };
{ data: TreeData; }

对应:

data
  name 根节点
  children
    - name 子节点A
      children
        - name 孙节点
    - name 子节点B

Syntax 完整示例

vis column
data
  - category A产品
    value 30
    group 线上
  - category B产品
    value 50
    group 线上
title 产品销量对比
axisXTitle 产品
axisYTitle 销量(万)
stack true
theme academy
style
  palette
    - #5B8FF9
    - #61DDAA

Markdown 语法

当输出为 Markdown 格式时,使用 GPT-Vis 作为 fenced code block 的语言标记,内容区写入完整的 Syntax 格式:

```GPT-Vis
vis line
data
  - time 2020
    value 100
```

格式:

```GPT-Vis
<完整的 Syntax 内容,首行 vis <chart-type>>
```

语法规则:

  • 语言标记固定为 GPT-Vis
  • 内容区使用 Syntax 格式规则编写,首行必须包含 vis <chart-type>(完整列表见上方支持的图表类型)
  • 代码块会被 Markdown 插件转换为 ,由浏览器端渲染

注意:Markdown 模式下,内容区必须写首行 vis <type>,与纯 Syntax 模式格式一致。

代码模式

根据目标框架生成完整可运行代码。

安装方式

NPM:

npm install @antv/gpt-vis
import { GPTVis } from '@antv/gpt-vis';

CDN:

<script src="https://unpkg.com/@antv/gpt-vis/dist/umd/index.min.js"></script>

CDN 引入后通过 GPTVis.GPTVis 访问主类。

HTML 完整示例

<html>
  <head>
    <script src="https://unpkg.com/@antv/gpt-vis/dist/umd/index.min.js"></script>
  </head>
  <body>
    <div id="container"></div>
    <script>
      const gptVis = new GPTVis.GPTVis({
        container: '#container',
        width: 600,
        height: 400,
      });

      gptVis.render(`
vis line
data
  - time 2020
    value 100
  - time 2021
    value 120
title 年度趋势
`);
    </script>
  </body>
</html>

图表类型配置

通用配置

所有图表均包含以下字段,后续各图表类型定义中省略这些字段。各小节标题即为 type 值(如 line、column),对应上方图表类型表中的 type 列。

{ type: string; title?: string; theme?: 'default' | 'light' | 'dark' | 'academy'; style?: { backgroundColor?: string; palette?: string[] } }

line / area

{ data: { time: string | number; value: number; group?: string }[]; axisXTitle?: string; axisYTitle?: string; stack?: boolean; style?: { lineWidth?: number } }

stack 仅 area 支持。

column / bar

{ data: { category: string; value: number; group?: string }[]; axisXTitle?: string; axisYTitle?: string; stack?: boolean; group?: boolean }

pie

value 不可使用百分比数字。

{ data: { category: string; value: number }[]; innerRadius?: number }

innerRadius 设为 0.6 变为环图。

scatter

{ data: { x: number; y: number; group?: string }[]; axisXTitle?: string; axisYTitle?: string }

dual-axes

{ categories: string[]; series: { type: 'line' | 'column'; data: number[]; axisYTitle?: string }[]; axisXTitle?: string; style?: { startAtZero?: boolean } }

histogram

{ data: number[]; binNumber?: number; axisXTitle?: string; axisYTitle?: string }

boxplot / violin

同一 category 需多条数据以展示分布。

{ data: { category: string; value: number; group?: string }[]; axisXTitle?: string; axisYTitle?: string; style?: { startAtZero?: boolean } }

radar

{ data: { name: string; value: number; group?: string }[]; align?: boolean }

align: 是否对齐各维度比例尺,默认 false(各轴独立缩放);true 时所有轴共享同一最大值,适合多系列绝对数值对比。

funnel

{ data: { category: string; value: number; }[]; }

waterfall

value 可为负数表示减少。palette 为色板数组,顺序为 [正值色, 负值色, 汇总色]。

{ data: { category: string; value: number }[]; axisXTitle?: string; axisYTitle?: string; style?: { palette?: string[] } }

liquid

percent 范围 0~1。

{ percent: number; shape?: 'rect' | 'circle' | 'pin' | 'triangle' }

word-cloud

{ data: { text: string; value: number; }[]; }

venn

交集用逗号分隔集合标识:sets: "A,B"。label 用于显示图表上对应集合的名称

{ data: { sets: string | string[]; value: number; label?: string }[] }

treemap

type TreeNode = { name: string; value: number; children?: TreeNode[] };
{ data: TreeNode[] }

sankey

{ data: { source: string; target: string; value: number }[]; nodeAlign?: 'left' | 'center' | 'right' | 'justify' }

flow-diagram / network-graph

source/target 引用节点的 name。

type GraphData = { nodes: { name: string }[]; edges: { source: string; target: string; name?: string }[] };

// flow-diagram
{ data: GraphData }

// network-graph
{ data: GraphData; layout?: 'force' | 'circular' | 'grid' | 'radial' | 'concentric' | 'dagre' }

mindmap / indented-tree / organization-chart

type TreeData = { name: string; children?: TreeData[] };

// mindmap
{ data: TreeData; direction?: 'H' | 'LR' | 'RL' }

// indented-tree
{ data: TreeData; direction?: 'LR' | 'RL' | 'H' }

// organization-chart
type OrganizationChartData = {
  name: string;
  description?: string;
  children?: OrganizationChartData[];
};

{ data: OrganizationChartData }

mindmap 默认 'H',indented-tree 默认 'LR'。

fishbone-diagram

type FishboneNode = { name: string; children?: FishboneNode[] };
{ data: FishboneNode; style?: { texture?: 'rough' | 'default' } }

texture: 'rough' 为手绘风格。

table

{ data: Record<string, string | number>[]; }

summary

summary 与其他图表类型完全不同:不使用 Syntax/JSON 配置,而是使用 T8 语法(Markdown + 语义标注)。

⚠️ 生成 summary 前必须:先读取 references/summary.md 获取 T8 语法规则、完整实体类型列表、属性字段定义、生成要求和示例,然后再生成内容。跳过此步骤将导致语法错误。

最佳实践

  1. 饼图分类不超过 5 个,超过建议合并为"其它"或改用条形图
  2. 不要用饼图展示趋势,不要用折线图展示无序分类
  3. 数值字段必须是数字类型,分类字段必须是文本类型
  4. 连续数值的分布(如薪资、成绩、年龄)必须用直方图(histogram)
  5. 多维数据字段映射:有两个分类维度时,x 轴维度写 time/category,另一个写 group
  6. 语法模式优先用 Syntax 格式(流式友好)
  7. 代码模式默认生成 HTML + CDN 方案(零安装),用户指定框架时再用 npm 方案

Icon Search

Use the icon HTTP API directly with curl.

API

Search Endpoint

  • Method: GET
  • URL: https://www.weavefox.cn/api/v1/infographic/icon
  • Query params:
    • text (required): search keyword, e.g. "data analysis"
    • topK (optional): number of icons to fetch (1-20), default 5

Example:

curl -sS -L --max-time 20 "https://www.weavefox.cn/api/v1/infographic/icon?text=document&topK=5"

Typical response:

{
  "success": true,
  "data": [
    "https://example.com/icon1.svg",
    "https://example.com/icon2.svg"
  ]
}

Retrieve SVG Content

curl -sS -L --max-time 20 "https://example.com/icon1.svg"

Workflow

  1. Determine the icon concept keyword (for example: security, document, data).
  2. Search icon URLs using the API endpoint.
  3. Use curl to fetch the SVG content of selected URLs.
  4. Use SVG directly in pages, diagrams, or infographic materials.

Notes

  • Use URL encoding for special characters in text.
  • topK range is 1–20; if omitted, the service returns up to 5 results.
  • For network issues, retry with a smaller topK or verify endpoint accessibility.

Sponsors

Details

Marketplace
antvis/chart-visualization-skills
Category
Community

More from antvis/chart-visualization-skills

All 11