
MCP Dashboards
Render interactive charts, dashboards and KPI widgets directly inside your AI conversation.
Add to your client
Copy the config for your MCP client and paste it into its config file.
claude mcp add dashboard -- npx -y mcp-dashboards --stdioPaste into ~/Library/Application Support/Claude/claude_desktop_config.json
{
"mcpServers": {
"mcp-dashboards": {
"command": "npx",
"args": [
"-y",
"mcp-dashboards",
"--stdio"
]
}
}
}Step-by-step guides: Add to Claude Desktop · Add to Cursor · Add to Windsurf
Before you start
- Node.js 18+
- An MCP Apps-compatible client (Claude Desktop, Claude Web, VS Code + Copilot, Goose, Postman, or MCPJam) for inline rendering; other clients get a browser preview link
About MCP Dashboards
MCP Dashboards turns data into interactive charts, dashboards, and KPI widgets rendered inline in any MCP Apps-compatible AI client (Claude Desktop, Claude Web, VS Code/Copilot, Goose, Postman, MCPJam). It exposes 31 chart-rendering tools, a dashboard composer, auto-detect and URL-fetch renderers, live-polling tools, visual catalogs for discovery, and chart-file management — with 21 themes, 8 typography options, 5 effect presets, and export to PPT/A4/PNG/CSV.
Tools & capabilities (38)
render_pie_chartPie/donut chart for composition — 'what makes up the whole?'
render_bar_chartBar chart for comparison — vertical, horizontal, stacked, drill-down.
render_line_chartLine/area chart for trends — smooth curves, gradient fills, time series.
render_scatter_chartScatter chart for relationships — per-point labels, annotations, quadrants.
render_candlestick_chartCandlestick chart for finance — OHLC data with volume bars.
render_radar_chartRadar chart for multi-axis comparison — skills, scores, product attributes.
render_treemap_chartTreemap for hierarchy — nested rectangles sized by value.
render_sankey_chartSankey diagram for flow — money, users, or resources between stages.
render_wordcloud_chartWord cloud for frequency — sized words from text analysis.
render_boxplot_chartBoxplot/violin for distribution — quartiles, outliers, density shapes.
render_live_chartLive real-time chart that auto-polls any MCP tool on a timer.
poll_httpData proxy that fetches JSON from any HTTP endpoint via secure presets or public URLs.
render_bullet_chartBullet chart for KPI vs target — 2-8 zone bands with labels.
render_lollipop_chartLollipop chart for ranking — clean dots with optional target markers.
render_dumbbell_chartDumbbell chart for gaps — before/after with scale labels and zone bands.
render_variance_chartVariance chart for budgets — actual vs budget, color-coded over/under.
render_funnel_chartFunnel chart for conversion — staged drop-off with percentages.
render_slope_chartSlope chart for change — ranking shifts between two periods.
render_waffle_chartWaffle chart for proportion — 10x10 grid showing composition.
render_sparkline_chartSparkline for compact trends — mini cards with change indicators.
render_radial_clusterRadial cluster for health checks — multi-metric ring gauges with status.
render_waterfall_chartWaterfall chart for cumulative impact — cascading bars.
render_heatmap_chartHeatmap for intensity — 2D grid with color mapping.
render_geo_chartGeo/choropleth map — color-coded countries by value.
render_bubble_mapBubble (pin) map — sized circles at lat/lng coordinates.
render_timeline_chartTimeline for progress — milestone tracker with status indicators.
render_hero_metricHero KPI widget — 11 variants (progress ring, gem, orb, NPS, etc.).
render_dashboardDashboard — KPI cards plus multiple charts in a responsive grid.
render_tableData table — sortable columns, striped rows, CSV export.
render_from_jsonAuto-detect — picks the best chart for any JSON data automatically.
render_from_urlFetches JSON from a URL and auto-visualizes it (SSRF-guarded).
render_catalogMaster catalog with 4 live preview tiles; click a tile and drill into a sub-catalog.
render_chart_catalogShows all 31 chart types with mini previews to find the right chart.
render_theme_catalogShows all 21 themes with color, typography, and effects previews.
render_hero_catalogShows all 11 hero metric variants for picking a KPI widget style.
render_effects_catalogShows all 5 effect presets, each applied to a real card.
list_chart_filesLists the chart preview HTML files written to the system temp folder.
delete_chart_filesDeletes chart preview HTML files (e.g. those older than a given age).
When to use it
- Compare metrics across periods or regions (e.g. Q1 vs Q2 revenue) as bar, line, or slope charts inline in chat
- Build a multi-widget dashboard with KPI cards and trend charts from a plain-English request
- Track live data (crypto prices, portfolio P/L, API metrics) that auto-updates on a timer via poll_http presets
- Visualize geography — color-coded choropleth maps or lat/lng bubble maps of users, sales, or traffic
- Show conversion funnels, financial candlesticks, distributions, or flows (Sankey) without leaving the conversation
- Export the rendered chart to PPT (16:9), paginated A4, PNG, or CSV for reports
Security notes
All processing happens locally; no data is collected or transmitted externally. External calls only happen when you explicitly provide a URL or preset (render_from_url / poll_http), and both refuse private, loopback, and link-local addresses (e.g. 192.168.*, 169.254.169.254) to block prompt-injection SSRF, plus per-host rate limiting. Credentials in env-var presets (POLL_PRESET_<NAME>_HEADERS) stay server-side. The browser preview server and optional HTTP transport bind to 127.0.0.1 by default; set MCP_HTTP_BIND_HOST=0.0.0.0 only on a trusted network. Chart HTML files live in the system temp folder and auto-delete after 7 days.
MCP Dashboards FAQ
Which AI clients support inline rendering?
Any MCP Apps-compatible client: Claude Desktop, Claude Web, VS Code (GitHub Copilot), Goose, Postman, and MCPJam. ChatGPT support is rolling out. Clients without MCP Apps still get a clickable http://localhost preview link plus a persistent file:// HTML artifact for each chart.
Do I need an API key?
No. The server requires no auth and all processing is local. API keys are only needed if you configure server-side live-polling presets (POLL_PRESET_<NAME>_URL / _HEADERS) to poll authenticated APIs — those credentials stay in env vars and never appear in the conversation.
Is my data sent anywhere?
No data is collected, transmitted, or stored externally. The only outbound calls happen when you explicitly provide a URL or preset for render_from_url / poll_http, and those are SSRF-guarded against private/loopback/link-local addresses and rate-limited per host.
How do I run it as a remote HTTP server?
Run `npx mcp-dashboards` and it starts a Streamable HTTP server on http://127.0.0.1:3001/mcp. It binds to localhost by default; set MCP_HTTP_BIND_HOST=0.0.0.0 and configure MCP_CORS_ALLOWED_ORIGINS to expose it on a trusted network.
Where are chart files stored?
Chart HTML files are written to <system temp>/mcp-dashboards/ and auto-deleted after 7 days (configurable via MCP_DASHBOARDS_RETAIN_DAYS; 0 disables cleanup). Use the chart's built-in PNG/PPT/A4 download button to save anything permanently, or set MCP_DASHBOARDS_DISABLE_PREVIEW to write no files at all.
Alternatives to MCP Dashboards
Compare all alternatives →Control Blender from Claude and other LLMs for prompt-assisted 3D modeling, scene creation, and asset generation.
Popular community server that feeds Figma layout data to coding agents via a Figma API token.
Generate beautiful, modern UI components from natural-language descriptions inside your IDE.
Compare MCP Dashboards with: