echarts-mcp
A Model Context Protocol server that renders Apache ECharts options to SVG, enabling AI agents to generate deterministic charts server-side without a browser.
README
echarts-mcp
A modern, open-source Apache ECharts MCP server. It exposes type-safe tools that let AI agents and chat UIs render rich, deterministic charts server-side — no browser, no native canvas dependency.
Built against the Model Context Protocol v2 SDK (@modelcontextprotocol/server@2) and
Apache ECharts 6.1.0 (pinned), supporting both stdio and Streamable HTTP transports.
Highlights
- SVG-first rendering —
echarts.init(null, theme, { renderer: "svg", ssr: true })+renderToSVGString()produces clean, scalable vector output with zero native dependencies and no DOM. Deterministic by default (animation: false). - Fully stateless — every tool result is self-contained (raw SVG + echoed option). No files, no session state, horizontally scalable over HTTP.
- Type-safe tool surface — JSON Schema derived from Zod v4 (
Standard Schema), so agents get precise, machine-readable input/output contracts. - Passthrough fidelity —
render_echartaccepts any ECharts v6 option; it does not restrict chart types (seelist_chart_typesfor discoverability only).
Tools
| Tool | Purpose |
|---|---|
render_echart |
Renders an ECharts option to SVG. Returns the raw SVG + echoed option. |
validate_echart_option |
Dry-run structural + semantic validation (pinned to echarts@6.1.0) without rendering. |
list_chart_types |
Lists common series types for agent discoverability (not an allowlist). |
render_echart inputs
| Param | Type | Default | Notes |
|---|---|---|---|
option |
object | required | Full ECharts v6 option (passthrough). |
renderer |
"svg" | "canvas" | "dual" |
"svg" |
canvas/raster ships in v1.1. |
width / height |
int | 800 / 500 | Render dimensions in pixels. |
devicePixelRatio |
number | 2 | Raster only (deferred). |
theme |
string | object | "light" |
Built-in name or inline theme JSON object. |
background |
string | "transparent" |
e.g. "#ffffff" or "transparent". |
animation |
boolean | false |
Emit an animated SVG when true. |
The result carries an MCP text content block (the raw SVG) plus structuredContent
with svg, width, height, renderer, theme, background, and version.
Quick start
Requires Node.js 20+ (developed on Node 24) and pnpm.
pnpm install
pnpm build # compiles src/ -> dist/
pnpm dev # run via tsx (stdio)
pnpm start # run the compiled dist (stdio)
Run the server over stdio (the default) from an MCP client by pointing it at dist/index.js.
Install from npm
npm install -g @ryuk3nd0/echarts-mcp
# then use it from any MCP client:
echarts-mcp
# or ad-hoc without a global install:
npx -y @ryuk3nd0/echarts-mcp
Streamable HTTP
MCP_TRANSPORT=streamable-http MCP_HTTP_PORT=3030 node dist/index.js
The MCP endpoint is then available at http://localhost:3030/mcp.
Configuration
Configuration is resolved in this precedence order (highest first):
- Environment variables
- A JSON config file (
MCP_CONFIGpath, or./config.jsonwhen present) - Built-in defaults
| Env var | Config key | Default |
|---|---|---|
MCP_TRANSPORT |
transport |
stdio |
MCP_HTTP_HOST |
http.host |
0.0.0.0 |
MCP_HTTP_PORT |
http.port |
3030 |
MCP_HTTP_PATH |
http.path |
/mcp |
MCP_API_TOKEN |
http.authToken |
null (disabled) |
See config.example.json for the full shape.
Authentication (Streamable HTTP only)
By default the HTTP endpoint is open (intended for trusted localhost). To protect it,
set MCP_API_TOKEN (or http.authToken in config.json). Clients must then send
Authorization: Bearer <token>; requests without a matching token receive 401.
Development
pnpm typecheck # tsc --noEmit
pnpm lint # eslint
pnpm test # vitest run
pnpm format # prettier
Docker
docker build -t echarts-mcp .
# stdio
docker run -i --rm echarts-mcp
# streamable HTTP
docker run -p 3030:3030 -e MCP_TRANSPORT=streamable-http echarts-mcp
Roadmap / deferred
- Canvas/PNG raster engine (v1.1): requires a DOM shim +
@napi-rs/canvas+echarts.setPlatformAPI({ createCanvas, loadImage, measureText })+ font registration. - Full OAuth 2.1 resource-server auth (optional bearer token today).
License
Apache-2.0. See LICENSE.
Integrations
Ready-to-paste client config snippets live in integrations/:
integrations/claude-desktop.json— Claude Desktop stdio entry.integrations/mcp.json— generic stdio entry (Claude Code, Cursor, and others).integrations/streamable-http.json— remote Streamable HTTP endpoint for gateways and hosted LLM tool integrations.
A machine-readable SKILL.md at the repo root describes the server for Agent Skills systems.
The same guidance is exposed to MCP clients via the server's instructions field.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
Neon Database
MCP server for interacting with Neon Management API and databases
E2B
Using MCP to run code via e2b.
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.