Input: Mermaid diagram syntax as a string
Processing:
- The diagram is passed to the Mermaid CLI via stdin
- The CLI validates the syntax and renders an SVG if valid
- Output and errors are captured from stdout/stderr
Output:
- Success: Text confirmation + rendered SVG as base64-encoded image
- Failure: Error message with details about the validation failure

Dependencies

External Libraries

@modelcontextprotocol/sdk: SDK for implementing Model Context Protocol
@mermaid-js/mermaid-cli: CLI tool for validating and rendering Mermaid diagrams
zod: Schema validation library for TypeScript

Development Dependencies

typescript: TypeScript compiler
eslint: Linting utility
prettier: Code formatting

API Specification

validateMermaid Tool

Purpose: Validates a Mermaid diagram and returns the rendered SVG if valid

Parameters:

diagram (string): The Mermaid diagram syntax to validate

Return Value:

Success:

{
  content: [
    { 
      type: "text", 
      text: "Mermaid diagram is valid" 
    },
    {
      type: "image", 
      data: string, // Base64-encoded SVG
      mimeType: "image/svg+xml"
    }
  ]
}

Failure:

{
  content: [
    { 
      type: "text", 
      text: "Mermaid diagram is invalid" 
    },
    {
      type: "text",
      text: string // Error message
    },
    {
      type: "text",
      text: string // Detailed error output (if available)
    }
  ]
}

Technical Decisions

MCP Integration: The project uses the Model Context Protocol to standardize the interface for AI tools, allowing seamless integration with compatible clients.
Child Process Approach: The implementation uses Node.js child processes to interact with the Mermaid CLI, which provides:
- Isolation between the main application and the rendering process
- Ability to capture detailed error information
- Proper handling of the rendering pipeline
Error Handling Strategy: The implementation uses a nested try-catch structure to:
- Distinguish between validation errors (invalid diagram syntax) and system errors
- Provide detailed error information to help users fix their diagrams
- Ensure the service remains stable even when processing invalid input
Simple Project Structure: The project uses a straightforward TypeScript project structure for:
- Easy maintenance and understanding
- Direct dependency management
- Simplified build process

Build and Execution

The application can be built and run using npm scripts:

# Install dependencies
npm install

# Build the application
npm run build

# Run locally (for development)
npx @modelcontextprotocol/inspector node dist/main.js

# Format code
npm run format

# Lint code
npm run lint

# Watch for changes (development)
npm run watch

The application runs as an MCP server that communicates via standard input/output, making it suitable for integration with MCP-compatible clients.

Release

To release a new version, the following steps in order:

npm run build
npm run bump
npm run changelog
npm publish --access public

d:

Categories

Current Category: mermaid

SCAST

mcp-mermaid-validator