# KyuRish/mcp-dashboards [Health: Active]

**Category:** 📊 Data Visualization  
**Repository:** https://github.com/KyuRish/mcp-dashboards  
**GitHub Stars:** 44  
**npm Downloads (last month):** 927  
**Views:** 5  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/kyurish-mcp-dashboards

## Description
45+ interactive chart types (bar, line, pie, candlestick, sankey, geo, radar, funnel, treemap, and more), dashboards with KPI cards, drill-down navigation, live API polling, 20 themes, and export to PNG/PPT/A4. Built on MCP Apps.

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `npx` (confidence: high):

```json
"mcpServers": {
  "mcp-dashboards": {
    "command": "npx",
    "args": ["-y","mcp-dashboards"]
  }
}
```

## Documentation

## What KyuRish/mcp-dashboards MCP server does

KyuRish/mcp-dashboards MCP server adds interactive data visualization to MCP-compatible AI conversations. It can render individual charts, multi-panel dashboards, KPI cards, sortable tables, and flowcharts as HTML views rather than static images. The available visualizations cover comparisons, trends, distributions, finance, geography, hierarchies, flows, rankings, proportions, timelines, and other common analytical tasks.

The server includes 32 chart-oriented tools covering more than 45 chart subtypes. It also provides catalogs for browsing chart types, themes, hero metric variants, and visual effects. An agent can select a specific renderer, use automatic JSON detection, or fetch JSON from a URL and choose a visualization.

## How it works

An MCP client sends a request describing the data and the desired result. The AI can select the appropriate rendering tool and return an interactive HTML visualization in the conversation. Users can hover for tooltips, select data points to provide context back to the AI, drill into supported charts with breadcrumbs, and zoom maps or heatmaps.

Flowcharts support editing actions such as dragging nodes, multi-select, renaming, minimap navigation, and step-through playback. Layout changes can be synchronized back to the AI. Live charts can poll an MCP tool on a timer, while the `poll_http` tool can retrieve JSON from public URLs or configured secure presets.

KyuRish/mcp-dashboards MCP server supports local stdio operation and Streamable HTTP operation. The documented HTTP startup command exposes the MCP endpoint at `http://127.0.0.1:3001/mcp` by default.

## Setup and configuration

The official npm package is `mcp-dashboards`. For Claude Desktop, add a server entry that runs `npx -y mcp-dashboards --stdio`. Claude Code and VS Code can use the equivalent `claude mcp add dashboard -- npx -y mcp-dashboards --stdio` command.

The README lists Claude Desktop, Claude Web, VS Code with GitHub Copilot, Goose, Postman, and MCPJam as supported MCP Apps-compatible clients. ChatGPT support is described as rolling out. The HTTP mode binds to localhost by default; network exposure or browser access from another origin requires additional configuration described by the project.

## Tools and capabilities

KyuRish/mcp-dashboards MCP server includes renderers for pie, bar, line, scatter, candlestick, radar, treemap, Sankey, word cloud, boxplot, bullet, funnel, slope, waffle, heatmap, geographic, bubble map, timeline, flowchart, hero metric, dashboard, and table views. Other tools cover radial clusters, waterfalls, sparklines, lollipop, dumbbell, variance, and live charts.

Charts can use 21 themes and configurable typography. Dashboards combine KPI cards and multiple charts in a responsive grid. Exports include PNG, PowerPoint in 16:9 format, paginated A4 output, and CSV where supported.

## Limitations and notes

The server requires an MCP Apps-compatible client for its inline interactive views. The documented default HTTP address is localhost, so it is not automatically a network-accessible service. Live polling and URL-based visualization depend on the availability and shape of the requested data source. The project identifies `mcp-dashboards` as its official npm package; similarly named packages should not be substituted.

_Full upstream README: https://allmcps.com/mcp/kyurish-mcp-dashboards/readme_