Data Visualization

Reviewed 2026-10-13dataPrompt / Rule# data-visualization # charts # matplotlib # seaborn # plotly

Anthropic's data-visualization skill: pick the right chart and write clean, accessible Python charts with matplotlib, seaborn and plotly.

Security Reviewed

Very low risk: a Markdown-only guide with plotting code examples and no network access.

What it looks like in Antigravity 2.0

Works (steps corrected)Installed and worked, but the agent skipped the Python library step, so the matplotlib code was never run.Run by us on 2026-10-13 · desktop app 2.21.1, Gemini 3.8 Flash (High), macOS

Sorted horizontal bar chart of 12 sales regions generated with the Data Visualization skill — What you get: the ranked bar chart it chose for 12 regions.

How to get this result, step by step

1Paste the install message into a new Antigravity conversation; reply "OK" when it lists the download.

Install the "Data Visualization" skill for Google Antigravity.
Follow the steps at https://www.antigravityskills.org/install/data-visualization.md (source: https://github.com/anthropics/knowledge-work-plugins/tree/main/data/skills/data-visualization).
Before running anything, show me what you will download and whether it includes scripts, and wait for my OK.
Install it into this project (.agents/skills/) unless I say global. When done, check that Antigravity lists it as an installed skill.

Antigravity 2.0 agent listing one SKILL.md file and no scripts before installing Data Visualization

2Ask for a chart in plain words — the agent loads the skill without you naming it.
```
Here is sales.csv. Rank the 12 regions by revenue and pick the right chart.
```
3It explains why it picked this chart type.

Full test log

Install method: Option 1 (pasted the agent prompt into a new conversation in an empty project folder).

The agent asked to read antigravityskills.org, github.com and raw.githubusercontent.com, then reported one file (SKILL.md), no scripts, target .agents/skills/data-visualization/SKILL.md, and waited for OK.
It cloned anthropics/knowledge-work-plugins and copied the skill folder (the Option 2 route). It did not run step 4 (pip install matplotlib seaborn pandas numpy plotly); matplotlib was still missing afterwards.
With the prompt below and a 12-row sales.csv we made, the agent read data-visualization/SKILL.md on its own (306 lines) and checked whether matplotlib, pandas and seaborn were installed.
It wrote plot_sales.py (matplotlib, horizontal bars) but, without matplotlib, produced the chart with a second, standard-library script that writes regional_revenue_ranking.svg.

What we corrected on this page

The agent skipped the pip install step when following Option 1. Run it yourself, or tell the agent to do it, if you want the matplotlib, seaborn or plotly code to run. Added to the step.
The check step said to run /skills; the 2.0 app has no such command. Changed to "What skills do you have?" for the app and /skills for the CLI.

Install Data Visualization in Google Antigravity

Antigravity loads skills from .agents/skills/ in your project or ~/.gemini/config/skills/ for all projects (official docs). Full guide

EasiestOption 1: paste this to your Antigravity agent

Install the "Data Visualization" skill for Google Antigravity.
Follow the steps at https://www.antigravityskills.org/install/data-visualization.md (source: https://github.com/anthropics/knowledge-work-plugins/tree/main/data/skills/data-visualization).
Before running anything, show me what you will download and whether it includes scripts, and wait for my OK.
Install it into this project (.agents/skills/) unless I say global. When done, check that Antigravity lists it as an installed skill.

The agent reads the install file, shows you what it will download and waits for your OK.

Option 2: install it yourself

1. Clone the source repository
```
git clone --depth 1 https://github.com/anthropics/knowledge-work-plugins /tmp/kwp
```
Clone the repository to a temporary folder so you can copy the skill folder out of it.
2. Copy into your workspace (this project only)
```
mkdir -p .agents/skills && cp -r /tmp/kwp/data/skills/data-visualization .agents/skills/data-visualization
```
Run this from your project root. Antigravity reads workspace skills from <project>/.agents/skills/<folder>/.
3. Or copy globally (all projects)
```
mkdir -p ~/.gemini/config/skills && cp -r /tmp/kwp/data/skills/data-visualization ~/.gemini/config/skills/data-visualization
```
Use this instead of the workspace step if you want the skill available in every project. Antigravity reads global skills from ~/.gemini/config/skills/<folder>/.
4. Install the Python libraries
```
pip install matplotlib seaborn pandas numpy plotly
```
The code patterns import these; plotly is only needed for interactive charts. In our Antigravity 2.0 test the agent skipped this step when following Option 1, so run it yourself (or ask the agent to) if you want the matplotlib, seaborn or plotly code to run.
5. Check that Antigravity sees it
Antigravity 2.0 app: start a new conversation and ask "What skills do you have?" (the app has no /skills command; typing / and the skill name also finds it, sometimes only after your first message). Antigravity CLI: run /skills; in a session that was already open, run /skills reload first. data-visualization should be listed.

The repository README does not mention Antigravity; these are Antigravity's standard skill folders. No API keys needed. Copy /tmp/kwp/data/skills/create-viz as well if you want the step-by-step charting workflow.

Using Claude Code, Cursor or Codex? See other assistants below.

What is Data Visualization?

data-visualization is Anthropic's chart skill for analysts and developers who ask an agent to turn data into charts: it tells the agent which chart fits the question, gives it ready matplotlib, seaborn and plotly code with a consistent style, and adds design and accessibility rules. It is one of the skills in the data plugin of anthropics/knowledge-work-plugins. What we counted at commit 8444efc: the skill is one file, SKILL.md (305 lines, 11,051 bytes), with no scripts. It contains a chart-selection table with 13 rows (trend, comparison, ranking, part-to-whole, composition over time, distribution, two correlation cases, geography, flow, network, performance vs target, multiple KPIs), 5 chart types to avoid or use with care (pie, 3D, dual-axis, many-category stacked bars, donut), 8 Python code blocks, and a 6-colour categorical palette. The code writes 5 PNG files (trend_chart, bar_chart, histogram, heatmap, small_multiples) and one HTML file (interactive_chart.html). The data plugin has 10 skill folders; note that its README lists skill names (data-exploration, data-validation, interactive-dashboard-builder) that do not exist as folders at this commit.

How to use Data Visualization?

Give the agent your data (a CSV, a DataFrame or query results) and say what you want to show. Because the frontmatter sets user-invocable: false, there is no /data-visualization command; the agent applies it from context. In Claude Code the plugin also gives you /create-viz for the full data-to-chart workflow. Ask for plotly when you want an interactive chart.

Key features of Data Visualization

✓Chart selection table by data relationship (trend, comparison, ranking, composition, distribution, correlation, geography, flow, KPIs)
✓Guidance on charts to avoid (pie with many categories, 3D, careless dual axes)
✓Code patterns for line, bar, histogram, heatmap and small-multiple charts
✓Number-formatting helpers and interactive plotly examples
✓Colourblind-friendly categorical, sequential and diverging palettes
✓Accessibility checklist (colour blindness, screen readers)

When to use Data Visualization

1Plot a metric trend by category over time
2Rank categories with a sorted horizontal bar chart
3Show a distribution with mean and median lines
4Build a correlation heatmap
5Create an interactive scatter chart with hover data

Installation & Setup Guide

Agent Rules / Prompt Setup

Follow these steps (taken from the project's own docs) to configure Data Visualization in your development environment.

Step 1: Add the marketplace (Claude Code)claude

claude plugin marketplace add anthropics/knowledge-work-plugins

Source: data/README.md, Installation.

Step 2: Install the data plugin (Claude Code)claude

claude plugin install data@knowledge-work-plugins

Source: data/README.md. Installs data-visualization together with create-viz, build-dashboard, sql-queries and the other data skills, plus the connector list in .mcp.json.

Step 3: Codex (skill-installer)codex

$skill-installer install https://github.com/anthropics/knowledge-work-plugins/tree/main/data/skills/data-visualization

Not documented by Anthropic for this repo. Source: openai/skills README ("provide the GitHub directory URL") and its skill-installer SKILL.md, which installs into $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex afterwards.

Step 4: Cursor: not documentedcursor

The repository README names Claude Cowork and Claude Code only, so we do not list a Cursor command.

Usage Example & Verification

Prompt / Request Example

1) README example (data/README.md, Ad-Hoc Analysis): "/analyze What was our monthly revenue trend for the past 12 months, broken down by product line?"
-> What happens (Claude Code with the data plugin): the agent writes the SQL, runs it if a warehouse MCP is connected, and draws a trend chart; data-visualization supplies the line-chart pattern and style.

2) Our example: "Here is sales.csv. Rank the 12 regions by revenue and pick the right chart."
-> What happens: following the selection table, the agent picks a sorted horizontal bar chart, uses the barh code pattern with value labels and saves bar_chart.png in the working folder.

3) Our example: "Make this chart readable for colour-blind readers."
-> What happens: the agent applies the accessibility checklist: different line styles or direct labels instead of colour alone, sns.color_palette("colorblind"), labelled axes with units.

Installation Confirmation Method

Ask "What skills do you have?" in Antigravity (Antigravity CLI: /skills) and look for data-visualization. Then ask "Which chart should I use to compare 3 categories over 12 months, and which charts should I avoid?" An answer that cites the selection table (line chart, no 3D, careful with dual axes) shows the skill is loaded.

Security Review

Static review

Reviewed Date

2026-10-13

Audited Ref / Commit

8444efcd48f7012f09797778a36a33e73d0861f4 (2026-10-01)

Reviewer

antigravityskills.org (static code review)

Very low risk: a Markdown-only guide with plotting code examples and no network access.

Full review: scope, findings, risks

Audit Scope:data/skills/data-visualization/SKILL.md (full), data/skills/create-viz/SKILL.md (header), data/README.md, data/.mcp.json, data/.claude-plugin/plugin.json, data/ file list; repository README.md and LICENSE, at commit 8444efc. The data/ folder is unchanged at HEAD 02ffa99 (checked 2026-10-13). Static review only; nothing was executed. The hands-on test on this page is separate.

Key Findings:Single Markdown file with Python code examples; no executable scripts in the folder. Examples write PNG files via plt.savefig in the working directory. No network calls, environment variables or API keys.

Identified Operational Risks & Permissions:Minimal. The agent runs Python you can see; output files are written to the working directory.

data/skills/data-visualization/ contains only SKILL.md (305 lines). Its Python snippets import matplotlib, seaborn, pandas, numpy and plotly and save charts with plt.savefig to files such as trend_chart.png, bar_chart.png and heatmap.png. There are no shell commands, HTTP requests, environment variables or credentials. The frontmatter sets user-invocable: false, so the skill is applied from context. Connectors for data warehouses belong to the wider data plugin, not this skill.

Operational Notes & Known Limitations

Known Limitations

•Python only: needs matplotlib, seaborn, pandas and numpy (plotly for interactive charts); it installs nothing for you
•Guidance plus code patterns, not a charting service; the agent must be able to run Python to render images
•user-invocable is false, so it activates from context rather than as a slash command
•Data warehouse access (Snowflake, BigQuery and others) comes from the data plugin's .mcp.json connectors (8 entries, two with empty URLs to fill in), not from this skill
•No Antigravity, Cursor or Codex instructions in the repository; the README targets Claude Cowork and Claude Code

Frequently Asked Questions

Is the Data Visualization skill safe?

Yes, in our static review it is a single SKILL.md with no scripts. Its code examples only plot data and save files such as trend_chart.png in the working folder; they make no network calls and need no API keys.

How do I install data-visualization in Antigravity?

Clone anthropics/knowledge-work-plugins and copy data/skills/data-visualization into .agents/skills/ (workspace) or ~/.gemini/config/skills/ (global), install the Python libraries, then check by asking "What skills do you have?" (Antigravity CLI: /skills).

Which libraries does it use?

matplotlib and seaborn for static charts, plotly (express and graph_objects) for interactive charts, with pandas and numpy for data handling.

Does it make dashboards?

No, it covers individual charts. The same data plugin has build-dashboard (HTML dashboards with Chart.js per the README) and create-viz for an end-to-end chart workflow.

Supported Platforms

antigravityclaudecodex

View Source on GitHub

Keep exploring this topic

Browse the full Antigravity skills directory for more source-linked skills with install steps and static safety reviews.

More data skills #data-visualization #charts #matplotlib #seaborn Use cases

Related Skills

📊

DuckDB CLI

@duckdb

Official DuckDB skills: query files and databases with the duckdb CLI, read CSV/Parquet/JSON, search DuckDB docs and install extensions.

Static reviewinstructions onlyinstalls packagesnetwork accesswrites files

duckdbsql

📊

Finance Tools (Financial Analysis)

@anthropics

Anthropic's financial-analysis plugin: 13 skills and 7 slash commands for DCF, comps, LBO and 3-statement models in Excel, spreadsheet audits and pitch-deck checks.

Static reviewruns scriptswrites filesnetwork access

financedcf

🎨