MCP Catalogs
Home

mcp-server-chart vs Embody

Side-by-side comparison to help you pick between these two MCP servers.

mcp-server-chart
by antvis
Embody
by dylanroscover
Stars★ 4,068★ 102
30d uses10,239
Score8447
Official
Categories
AI / LLM ToolsDeveloper ToolsProductivity
Developer ToolsMediaAI / LLM Tools
LanguageTypeScriptPython
Last committhis monththis month

mcp-server-chart · Summary

A TypeScript MCP server for generating 26+ visualization charts using AntV, supporting multiple chart types and deployment options.

Embody · Summary

MCP server for TouchDesigner that lets AI assistants build, wire, and debug networks with natural language commands.

mcp-server-chart · Use cases

  • Data analysts creating visual reports from datasets
  • AI assistants generating custom charts based on user requests
  • Web applications embedding visualization capabilities via HTTP API

Embody · Use cases

  • AI-assisted visual programming in TouchDesigner without manual operator creation
  • Version control and branching of complex visual projects through diffable network files
  • Debugging TouchDesigner networks through natural language commands to AI assistants

mcp-server-chart · Install

Installation

Install globally:

npm install -g @antv/mcp-server-chart

For Desktop Apps (e.g., Claude Desktop, VSCode):

{
  "mcpServers": {
    "mcp-server-chart": {
      "command": "npx",
      "args": ["-y", "@antv/mcp-server-chart"]
    }
  }
}

For Windows:

{
  "mcpServers": {
    "mcp-server-chart": {
      "command": "cmd",
      "args": ["/c", "npx", "-y", "@antv/mcp-server-chart"]
    }
  }
}

Embody · Install

Installation

  1. Download the Embody .tox from the [release page](https://github.com/dylanroscover/Embody/releases)
  2. Drag it into your TouchDesigner project
  3. Tag operators by selecting any COMP or DAT and pressing lctrl twice
  4. Enable the Envoy MCP server by toggling the Envoyenable parameter on the Embody COMP

The server starts on localhost:9870 and auto-creates a .mcp.json in your repo root.

For Claude Code integration, the .mcp.json should look like:

{
  "mcpServers": {
    "envoy": {
      "type": "http",
      "url": "http://localhost:9870/mcp"
    }
  }
}
Comparison generated from public README + GitHub signals. Last updated automatically.