data-viz-agent — Claude Code skill: charts, accessibility, install
data-viz-agent builds charts that hold up: picks the chart type from the data relationship, orders encodings by perceptual accuracy, and checks contrast and keyboard access.
MIT 204 installs ★ 35 updated 2026-10-21
npx skills add podo/design-agent-skills --skill data-viz-agent -a claude-codeWhat data-viz-agent does
The skill works like a visualization engineer on the team. It reads the relationship in the data, chooses a chart type for it, and assigns values to channels in order of how precisely people read them: position first, then length, colour last. On the web it works with D3.js, Chart.js, Plotly or Observable; for analysis, Matplotlib, Seaborn or Altair. Before handing a chart back it checks WCAG AA contrast, alt text, keyboard operation and a palette that survives color blindness.
The install leaves a pointer behind. The actual definition is a Claude subagent file, and the skills CLI registers it under the name data-visualization, so two entries sit side by side instead of one being replaced. Trigger phrases such as “Chart.js” or “Plotly” bring it in.
One thing to watch: designer-skills also ships a skill called data-visualization. Install both and whichever lands last overwrites the other. Single-file Tufte-style D3 work belongs to claud3 or d3js-skill, AntV/G2 charts to antvis-chart-skills, and dataset write-ups to data-analysis-skill.
What it helps with
- pick a chart type that matches the data relationship
- build a Chart.js or Plotly chart without blowing contrast
- test a palette for colorblind readability
- add keyboard control and alt text to an interactive chart
- decide between Matplotlib and D3.js for a given job
Example prompts for the agent
- Chart the monthly totals from
sales.csvas bars, and keep the palette readable for colorblind readers. - Rebuild this chart in Chart.js with WCAG AA contrast and keyboard support.
- Which chart type fits price against demand? Draw it in Plotly.
How to install data-viz-agent
1Skills CLI (Vercel)
npx skills add podo/design-agent-skills --skill data-viz-agent -a claude-code2Manually: put the files in place
~/.claude/skills/data-viz-agent/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/data-viz-agent/
1Skills CLI (Vercel)
npx skills add podo/design-agent-skills --skill data-viz-agent -a codex2Manually: put the files in place
.agents/skills/data-viz-agent/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add podo/design-agent-skills --skill data-viz-agent -a cursor2Manually: put the files in place
.cursor/skills/data-viz-agent/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install podo/design-agent-skills skills/data-viz-agent2Manually: put the files in place
.github/skills/data-viz-agent/SKILL.md1Skills CLI (Vercel)
npx skills add podo/design-agent-skills --skill data-viz-agent -a gemini-cli2Platform CLI
gemini skills install https://github.com/podo/design-agent-skills --path skills/data-viz-agentManually: put the files in place
.devin/skills/data-viz-agent/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add podo/design-agent-skills --skill data-viz-agent -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about data-viz-agent
How do I install data-viz-agent in Claude Code?
Run npx skills add podo/design-agent-skills --skill data-viz-agent -a claude-code; the skill file lands at ~/.claude/skills/data-viz-agent/SKILL.md.
What does data-viz-agent do?
It picks a chart type for the data, maps values to perceptual channels by accuracy, and checks the result for WCAG AA contrast, alt text, keyboard operation and colorblind safety.
Why does it install under a different name?
The subagent definition is registered as data-visualization, while data-viz-agent stays as a pointer next to it. designer-skills uses the same data-visualization name, so installing both makes the later one overwrite the earlier.
Source text
SKILL.md · MIT
--- name: data-viz-agent description: | Data visualization engineer agent using D3.js, Chart.js, Matplotlib, and Plotly. Chooses visual encodings by perceptual accuracy, enforces WCAG AA contrast, colorblind-safe palettes, and keyboard and screen-reader support, and selects the right chart type for the data relationship. Upstream is a Claude Code subagent definition installed here as a skill. triggers: - "data visualization agent" - "chart agent" - "visualization engineer" - "Chart.js" - "Plotly" - "Matplotlib" - "data chart" das: type: skill category: data-visualization upstream: "https://github.com/rohitg00/awesome-claude-code-toolkit" upstream_path: "agents/data-ai/data-visualization.md" version: latest install: true --- # data-viz-agent > Catalogue stub — full skill: [rohitg00/awesome-claude-code-toolkit](https://github.com/rohitg00/awesome-claude-code-toolkit/blob/main/agents/data-ai/data-visualization.md) <!-- das:install:start — generated from stub.yaml by `npm run generate`; edit stub.yaml, not this block --> ## Decision tree 1. **Is the full skill already installed?** Upstream names this skill `data-visualization`, so it installs next to this pointer (the pointer is not replaced). - Global: `f=~/.agents/skills/data-visualization/SKILL.md; [ -f "$f" ] && ! grep -q "^das:" "$f" && echo "installed" || echo "pointer"` - Project: `f=.agents/skills/data-visualization/SKILL.md; [ -f "$f" ] && ! grep -q "^das:" "$f" && echo "installed" || echo "pointer"` - `installed` → the skill is installed as `data-visualization`, invoke it and proceed - `pointer` → go to step 2 2. **Detect scope, then install:** ```bash [ -e ~/.agents/skills/data-viz-agent ] && echo "global" || echo "project" ``` **Global:** ```bash npx skills add https://raw.githubusercontent.com/rohitg00/awesome-claude-code-toolkit/main/agents/data-ai/data-visualization.md -g -y ``` **Project:** ```bash npx skills add https://raw.githubusercontent.com/rohitg00/awesome-claude-code-toolkit/main/agents/data-ai/data-visualization.md -y ``` > **Claude Code:** send a command as a chat message starting with `!` to run it without leaving the conversation. **Heads-up** - **License:** Apache-2.0 - **Name collision:** `data-visualization` is also written by `designer-skills` (its install or its catalogue pointer); installing both overwrites the earlier one. <!-- das:install:end --> ## Invoke after install - Skill name: `data-visualization` - Trigger phrases: "data visualization agent", "Chart.js", "Plotly", "Matplotlib", "data chart" ## What it does A data visualization engineering persona that works across D3.js, Chart.js, Plotly, and Observable on the web and Matplotlib, Seaborn, and Altair for analysis. It maps data dimensions to perceptual channels in order of accuracy (position for quantitative comparison, then length, then color), applies accessibility standards (WCAG AA contrast, alternative text, keyboard-navigable interactions, colorblind-safe palettes), picks the library that fits the job, and verifies the result before handing it back. The upstream file is a Claude Code subagent definition (its frontmatter declares tools and a model) rather than a SKILL.md; the skills CLI installs it from the raw file URL as a skill named `data-visualization`. ## When NOT to use - Tufte-style single-file D3 charts — use `claud3` or `d3js-skill` - AntV/G2 charts — use `antvis-chart-skills` - Dataset analysis and report writing rather than chart building — use `data-analysis-skill`
© podo · MIT · podo/design-agent-skills