Mastergo Magic MCP Server

MasterGo design files retrieve DSL data for your AI models via MCP protocol.

Data last scanned last week · Reviewed last month

408 tools★ 274TypeScript

Overview

MasterGo Magic MCP provides a direct connection between MasterGo design files and AI models, allowing for the retrieval of DSL data. This integration supports design-to-code tasks, enabling AI assistants to extract SVG components or restore designs directly from design URLs. The server operates as a standalone service runnable via npx. Users must provide a personal access token generated through their MasterGo security settings to authenticate the connection. The tool requires a Team Edition account or higher, and files must be hosted in Team Projects rather than personal drafts.

Our verdict

With a security score of 40/100, this server is classified as high-risk. The lack of a declared software license introduces significant legal ambiguity regarding its use. Furthermore, because it is a community-maintained project rather than an official vendor release, it has not undergone formal security audits for authentication or read-only protocols. It is recommended only for users who can independently vet the source code before deployment.

Supports direct extraction of design DSL data
Minimal supply-chain surface with five dependencies
Requires only Node.js environment via npx
Documented support for custom proxy and header configuration

Unclear legal status due to missing software license
Not maintained by the official MasterGo organization
Authentication methods remain unreviewed for security
Requires paid Team Edition MasterGo account

Setup

Claude Desktop

{
  "mcpServers": {
    "mastergo-magic-mcp": {
      "args": [
        "-y",
        "@mastergo/magic-mcp"
      ],
      "command": "npx"
    }
  }
}

Paste into ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows), then restart Claude Desktop. Some servers require API keys or tokens as environment variables — check the project's README and add an "env" object if needed.

No read-only mode — running this grants full read/write access.
No built-in authentication — anyone with access to this config can invoke every tool it exposes.

Tools

Tool	Description	Risk
mcp__getDesignSections	Retrieves section list and per-section DSL from MasterGo design files.	low
mcp__getDsl	Retrieves DSL (Domain Specific Language) data from MasterGo design files.	low
mcp__getDesignSvgs	Retrieves SVG data from MasterGo design files.	low
mcp__getDesignTexts	Retrieves text content from MasterGo design files.	low
mcp__extractSvg	Extracts SVG content from MasterGo design files.	low
mcp__getMeta	Retrieves metadata information.	low
mcp__getComponentLink	Retrieves component documentation from links.	low
mcp__getComponentWorkflow	Provides structured component development workflow for Vue and React components, generating workflow files and component specifications.	low

Compatibility

Client	Local	Docker	Remote	Read-only
ChatGPT
Claude Desktop
Cursor
VS Code
Windsurf

Frequently asked questions

›What are the account and file requirements for this server?

You must use a MasterGo Team Edition account or higher, as personal free editions are not supported. Additionally, design files must be stored in Team Projects rather than the draft box to be accessible.

›How do I obtain an API token?

Log in to MasterGo, navigate to your personal settings, select the Security Settings tab, and generate a new personal access token.

›Why might my AI client fail to list or use the server's tools?

Some clients, such as Grok Build, may silently skip tools if the naming prefix causes multiple double-underscore separators. If this happens, use the --no-prefix flag or set the MG_NO_PREFIX environment variable to 1.

›Can I influence the format of the output data returned by the server?

Yes, you can set a default format of json, yaml, or tree using the --format command-line argument or the DEFAULT_FORMAT environment variable. You can also specify the format per tool call if the AI model supports passing parameters.

›Are there any external dependencies required to run the server?

No external dependencies are required. The server runs directly via npx and only requires a Node.js environment.

›How can I provide custom headers to the server requests?

You can provide custom headers via the --header command-line argument or by setting the MG_EXTRA_HEADERS environment variable as a JSON object. Note that custom headers will override default headers like Content-Type and the auth token, so ensure you match the keys correctly.