home / mcp / mcp server chart mcp server
Provides a TypeScript-based MCP server that generates AntV charts and analyzes data.
Configuration
View docs{
"mcpServers": {
"antvis-mcp-server-chart": {
"command": "npx",
"args": [
"-y",
"@antv/mcp-server-chart"
],
"env": {
"SERVICE_ID": "your-service-id-123",
"DISABLED_TOOLS": "generate_fishbone_diagram,generate_mind_map",
"VIS_REQUEST_SERVER": "https://your-server.com/api/chart"
}
}
}
}You can deploy and run a TypeScript-based MCP server that generates AntV charts and performs data analysis. This server exposes a range of chart generation capabilities through MCP tooling, enabling you to produce visualizations directly from your data pipelines or AI-assisted workflows.
You interact with the MCP server by configuring it in your MCP client to run locally or remotely. You can choose between transport methods such as server-sent events (SSE) or a streamable transport depending on your client’s capabilities. Once configured, you can request chart generation or data analysis tasks through the MCP client, and the server returns chart outputs or data-driven insights. If you are using an AI IDE with skills support, you can leverage specialized chart-visualization skills to automatically select the best chart type and generate visualizations.
Practical usage patterns include setting up the server in your development environment, connecting your MCP-capable client, and selecting the chart type you need. You can also opt for private deployment by pointing the server to your own chart generation service endpoint if you require a self-contained or restricted environment.
Prerequisites: ensure you have Node.js and npm installed on your system. You may also install locally or run via a package runner to fetch the MCP server executable.
Install, build, and start the server with these steps if you are running locally on your machine:
# Install dependencies and build the server
npm install
npm run build
# Start the MCP server with a standard transport (stdio) for local use
npm run start
# Or start directly with a specific transport if you prefer SSE or streamable
node build/index.js -t sseEnvironment variables you may configure include SERVICE_ID, VIS_REQUEST_SERVER, and DISABLED_TOOLS to tailor the server behavior and available chart tools.
Private deployment lets you point the server to a private chart generation service. You can provide your own endpoint by setting VIS_REQUEST_SERVER in the server configuration and restarting your client.
If you want to generate mapping records or visualize map data, you can enable the chart generation flow and view results on your own records page after generating a service identifier.
Tool filtering allows you to disable specific chart generation tools by configuring DISABLED_TOOLS so you can constrain available functionality to match client capabilities or requirements.
For development, you can build and run the server locally using standard npm commands, or run the final build with a direct Node command for SSE or streamable transports.
The project is released under MIT, allowing you to adopt, modify, and deploy in your environment.
You can customize the server’s behavior with the following environment variables.
To use a private chart generation service, configure VIS_REQUEST_SERVER to point to your own HTTP endpoint and restart your client so requests are routed to your service.
{
"mcpServers": {
"mcp-server-chart": {
"command": "npx",
"args": ["-y", "@antv/mcp-server-chart"],
"env": {
"VIS_REQUEST_SERVER": "<YOUR_VIS_REQUEST_SERVER>"
}
}
}
}If you want to view your chart generation records, you can obtain a service identifier and configure it in your MCP client. Follow the process to generate a personal service identifier and include it in the MCP server configuration under SERVICE_ID.
Use the DISABLED_TOOLS environment variable to disable specific chart generation tools when needed.
Development steps include installing dependencies, building, and starting the server. You can run the server with SSE or streamable transport.
MIT license applies to this project.
Generate an area chart to display trends over a continuous variable and observe overall data patterns.
Generate a bar chart to compare values across different categories, suitable for horizontal comparisons.
Generate a boxplot to visualize data distribution, including median, quartiles, and outliers.
Generate a column chart to compare values across categories in a vertical layout.
Generate a district map to show administrative divisions and data distribution.
Generate a dual-axes chart to display relationships between two variables with different units.
Generate a fishbone (Ishikawa) diagram to identify root causes of a problem.
Generate a flowchart to visualize steps and sequences in a process.
Generate a funnel chart to illustrate data loss across stages.
Generate a histogram to display data distribution by counting observations in intervals.
Generate a line chart to show data trends over time or another continuous variable.
Generate a liquid chart to visualize proportions using water-fill representations.
Generate a mind-map to display hierarchical information and thought structure.
Generate a network graph to show relationships and connections between nodes.
Generate an organizational chart to display structure and personnel relationships.
Generate a path-map to visualize route planning results for points of interest.
Generate a pie chart to show data proportions as segments of a circle.
Generate a pin-map to display the geographic distribution of points of interest.
Generate a radar chart to display multi-dimensional data across multiple axes.
Generate a sankey chart to illustrate data flow and volume between nodes.
Generate a scatter plot to show relationships between two variables.
Generate a treemap to visualize hierarchical data with area-encoded values.
Generate a Venn diagram to show relationships between sets.
Generate a violin plot combining density and distribution information.
Generate a word cloud to display word frequency with font size indicating prevalence.
Generate a spreadsheet or pivot table for tabular data, rendering as cross-tabulation when rows/values are provided.