mcp-mermaid

MCP Mermaid is an MCP server that generates Mermaid diagrams and charts using AI assistance. It leverages the Mermaid syntax and output options, and can export results in formats like mermaid text, SVG, PNG, or base64 representations, enabling easy integration with AI-assisted workflows. The server is accessible via standard MCP transports (SSE or Streamable) or through local development builds, and can be invoked by an MCP-enabled client or agent to request diagram generation or diagram-based insights. To get started, configure the MCP server in your environment and point your MCP client to the mcp-mermaid server entry, then send requests that specify Mermaid diagram data and optional styling preferences. The server supports full Mermaid syntax and can adapt outputs based on the request details (diagram type, theme, background color, and export format).

Prerequisites:

• Node.js and npm installed on your system
• Internet access to fetch the mcp-mermaid package from npm

Option A: Global installation (recommended for quick testing)

1. Install globally: npm install -g mcp-mermaid
2. Run the server via MCP transport (SSE by default) or use your MCP client to connect to it.

Option B: Local development and build

1. Clone the repository or download the package sources.
2. Install dependencies: npm install
3. Build the project (if applicable): npm run build
4. Run the server locally or via npm scripts (see package.json for available scripts).

Using either option, ensure your MCP client is configured to connect to the server with the appropriate transport (SSE or Streamable) and endpoint as described in the documentation.

Tips and common considerations:

• The server exposes multiple transport options; choose SSE for simple streaming or Streamable for more flexible HTTP interactions.
• When using local development, you can start with npm run start:sse or npm run start:streamable to verify the server is up before integrating into your MCP workflow.
• If you encounter port conflicts, adjust the default ports via the CLI options described in the README (transport, port, endpoint).
• Ensure the environment where the server runs has network access to fetch dependencies and any AI model resources required for Mermaid generation.
• The npm package name is mcp-mermaid; the npm_package field below reflects this value.