Datawrapper — Claude Code MCP server: what it does, how to install
Creates and customizes Datawrapper charts through chat prompts. Useful for plotting raw data, tweaking visual palettes, and publishing web graphics.
claude mcp add datawrapper-mcp -e DATAWRAPPER_ACCESS_TOKEN=<DATAWRAPPER_ACCESS_TOKEN> -- uvx datawrapper-mcpWhat Datawrapper does
This server connects your assistant to Datawrapper using an underlying Python library. It exposes eight tools to build, inspect, restyle, delete, and publish graphics. Supported formats include line, column, and scatter charts.
Installing through marketplace plugin pickers is risky right now because those flows cannot collect environment variables. Because of this bug, defining DATAWRAPPER_ACCESS_TOKEN manually in your configuration file or CLI command remains mandatory. You also need Python 3.10 or later on your machine.
You can set DATAWRAPPER_MCP_INSTRUCTIONS with your own style rules. The model folds those guidelines into its context before making API calls.
What it helps with
- build Datawrapper charts from raw data points inside chat
- modify colors and layout settings on existing graphics
- publish visualizations and retrieve live editor links
- render and export finished charts to PNG images
Example prompts for the agent
- Create a line chart showing annual revenue: 2021: 12.4, 2022: 15.1, 2023: 18.0.
- Change the line color to blue and publish this chart.
- Export chart
abc123as a PNG file.
How to install Datawrapper
Platform CLI
claude mcp add datawrapper-mcp -e DATAWRAPPER_ACCESS_TOKEN=<DATAWRAPPER_ACCESS_TOKEN> -- uvx datawrapper-mcpReplace <…> with your own values. uvx runs the command named after the package; if the README names another one, use uvx --from datawrapper-mcp <command>
Platform CLI
codex mcp add datawrapper --env DATAWRAPPER_ACCESS_TOKEN=<DATAWRAPPER_ACCESS_TOKEN> -- uvx datawrapper-mcpReplace <…> with your own values. uvx runs the command named after the package; if the README names another one, use uvx --from datawrapper-mcp <command>
Config file
{
"mcpServers": {
"datawrapper": {
"command": "uvx",
"args": [
"datawrapper-mcp"
],
"env": {
"DATAWRAPPER_ACCESS_TOKEN": "<DATAWRAPPER_ACCESS_TOKEN>"
}
}
}
}Replace <…> with your own values. uvx runs the command named after the package; if the README names another one, use uvx --from datawrapper-mcp <command>
Config file
{
"servers": {
"datawrapper": {
"type": "stdio",
"command": "uvx",
"args": [
"datawrapper-mcp"
],
"env": {
"DATAWRAPPER_ACCESS_TOKEN": "<DATAWRAPPER_ACCESS_TOKEN>"
}
}
}
}Replace <…> with your own values. uvx runs the command named after the package; if the README names another one, use uvx --from datawrapper-mcp <command>
Platform CLI
gemini mcp add -e DATAWRAPPER_ACCESS_TOKEN=<DATAWRAPPER_ACCESS_TOKEN> datawrapper uvx datawrapper-mcpReplace <…> with your own values. uvx runs the command named after the package; if the README names another one, use uvx --from datawrapper-mcp <command>
Step by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more
Questions about Datawrapper
How to install Datawrapper in Claude Code?
Run claude mcp add datawrapper-mcp -e DATAWRAPPER_ACCESS_TOKEN=<DATAWRAPPER_ACCESS_TOKEN> -- uvx datawrapper-mcp in your terminal. You must generate the token in your Datawrapper account settings first.
What chart types does Datawrapper support?
It handles bar, line, area, arrow, column, multiple column, scatter, and stacked bar charts using create_chart and update_chart.
What are the prerequisites for this MCP server?
You need a Datawrapper account with an API token, the uvx command, and Python 3.10 or higher.
