data-visualization Skill (Anthropic) for Claude Code: Charts

Helps Claude choose the right chart type and write accessible Python charts with matplotlib, seaborn or Plotly.

Installs: 14.1K; on skills.sh
Stars: 34.4K; on GitHub
Updated: Nov 2026; last repo commit

The skill gives Claude a chart selection guide, Python code patterns, design rules and an accessibility checklist. It maps what you want to show, such as a trend, a ranking or a distribution, to a chart type, and warns against pie, 3D and dual-axis charts in most cases.

It is marked user-invocable: false, so it is not a slash command. Claude draws on it when you ask for a chart. The code saves PNG or HTML files, and Claude needs matplotlib, seaborn, pandas or Plotly available in your Python environment.

Good for

Analysts turning a dataset into a chart for a report
Data scientists who want publication-quality figures
Anyone who must make charts that color-blind readers can read

Install data-visualization

Claude Code

Copy this prompt.

Install the data-visualization skill in Claude Code. Run this command:

claude plugin marketplace add anthropics/knowledge-work-plugins
claude plugin install data@knowledge-work-plugins

Then check that it loaded and tell me how to use it.

2
Paste it into Claude Code. It shows the change and asks before running it.

Requires Claude Code.

The plugin installs the whole data plugin. The skills CLI route installs only this skill, with no connectors.

Built from the package data we hold. Not yet tested by us. Source.

What's in the skill

When it loads

Create effective data visualizations with Python (matplotlib, seaborn, plotly). Use when building charts, choosing the right chart type for a dataset, creating publication-quality figures, or applying design principles like accessibility and color theory.

Sections in SKILL.md

Chart Selection Guide
Choose by Data Relationship
When NOT to Use Certain Charts
Python Visualization Code Patterns
Setup and Style
Line Chart (Time Series)
Bar Chart (Comparison)
Histogram (Distribution)
Heatmap
Small Multiples
Number Formatting Helpers
Interactive Charts with Plotly
Design Principles
Color
Typography
Layout

Key facts

Author: Anthropic
Repository: anthropics/knowledge-work-plugins
Installs on skills.sh: 14,050 as of 2026-11-02
GitHub stars: 34,361 for the whole anthropics/knowledge-work-plugins repo
License: Apache-2.0 (repo)
Last repo commit: 2026-11-01
What is inside: 1 file: SKILL.md. It bundles no scripts. The data plugin around it has 10 skills and a .mcp.json.
skills.sh audits: Gen Agent Trust Hub: pass; Socket: pass; Snyk: pass
Part of: The data plugin, version 1.1.0, by Anthropic per plugin.json
Invocation: user-invocable: false, so Claude uses it automatically rather than as a slash command

What it includes

A table from data relationship to best chart
Maps trend, comparison, ranking, composition, distribution, correlation, geography, flow and KPIs to a best chart and alternatives.
Try: “Which chart should I use to compare revenue across 12 regions?”
Chart selection guide
Pie, 3D, dual-axis, crowded stacked bars and donuts
Says to avoid pie charts unless there are fewer than 6 categories, never to use 3D charts, to use dual axes cautiously and to treat donuts as a single-KPI display at most.
When not to use a chart
Style setup, line, bar, histogram, heatmap and small multiples
Matplotlib and seaborn examples that call plt.savefig to write PNG files, plus number-format helpers for currency and large numbers.
Write · Python patterns
Plotly line and scatter charts saved as HTML
Plotly Express examples that call fig.write_html and fig.show.
Write · Interactive charts
Color, typography, layout and accuracy rules
Use color to encode data, start bar charts at zero, sort categories by value, and title charts with the insight.
Design principles
Color blindness, screen readers and a pre-share checklist
Do not rely on color alone, add alt text and a data table, and use the colorblind palette from seaborn.
Accessibility checklist

From the official docs

Permissions

Low risk

The skill alone is instructions, and its examples write chart files. The data plugin route also adds a .mcp.json with 8 data connectors.

What it can do

Read your data
Create and change things
Delete things
Read-only mode

Watch out for

The plugin route brings 8 data connectors
The data plugin's .mcp.json lists remote http connections to data tools. Use the skills CLI route if you want just this skill. github.com
Example code writes files
The examples call plt.savefig and fig.write_html, which write files in the working folder. github.com

Security advisories

No published security advisories as of Nov 2, 2026.

Technical details

Turning on read-only	No read-only switch for the skill. For the plugin, you choose which connectors to sign in to.
Talks to	Plugin route: .mcp.json lists http endpoints at bigquery.googleapis.com, app.hex.tech, mcp.amplitude.com, mcp.eu.amplitude.com, mcp.atlassian.com and api.definite.app, and two entries (snowflake, databricks) have empty URLs
Sign-in and keys	Connectors in the plugin route use their own sign-in. SKILL.md mentions no credentials.
Scopes	Claude writes chart images and HTML files into your working folder when you ask Plugin route: data/.mcp.json lists connectors for Snowflake, Databricks, BigQuery, Hex, Amplitude (two regions), Atlassian and Definite

Questions

What does the data-visualization skill do?

It guides Claude to pick a suitable chart and write clean Python code for it. It covers matplotlib, seaborn and Plotly, plus color, labeling and accessibility rules.

How do I install it?

Use the install box above. The README documents installing Anthropic’s data plugin, which includes this skill and nine others. The skills CLI route installs only this skill.

Can I call it with a slash command?

No. Its frontmatter says user-invocable: false, so Claude uses it automatically when you ask for a chart.

Does it need Python libraries?

The code examples import matplotlib, seaborn, pandas, numpy and Plotly. The skill does not install them, so they must already be in your environment.

What else comes with the data plugin?

The plugin has 10 skills, including SQL, statistics and dashboard skills, and a .mcp.json that lists 8 connectors such as Snowflake, BigQuery, Databricks and Hex. The skills CLI route installs none of the connectors.

Is it safe to install?

The skill alone bundles no scripts and only holds instructions and code examples. The examples write image or HTML files. The plugin route also brings the connector list described in the permissions section.

Information

updated Nov 2, 2026

Repository: anthropics/knowledge-work-plugins
Folder: data/skills/data-visualization

anthropics/knowledge-work-plugins skills.sh

Sources

Docs, issue trackers and discussions this page is built from.

Show all 6

Using skills in Claude · support.claude.com
knowledge-work-plugins README · github.com
data plugin .mcp.json · github.com
skills CLI commands · github.com
data-visualization SKILL.md · github.com
data plugin.json · github.com