home / mcp / mendix mcp server

Mendix MCP Server

Provides offline and cloud-assisted access to Mendix project data for AI agents, exposing Microflows and Domain Models as AI-friendly JSON.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "egorsel-mendix-local-mcp": {
      "command": "node",
      "args": [
        "/absolute/path/to/mendix-local-mcp/build/server.js"
      ],
      "env": {
        "MENDIX_TOKEN": "YOUR_TOKEN_PLACEHOLDER",
        "MENDIX_USERNAME": "YOUR_EMAIL_PLACEHOLDER"
      }
    }
  }
}

The Mendix Local MCP Server is a bridge that lets Mendix projects be inspected and leveraged by AI tools. It provides offline, fast access to project data and, when needed, cloud-backed precision, so you can work with Microflows, Domain Models, and other Mendix structures in AI workflows.

How to use

Start the server in your local environment and connect your MCP client to it. You can operate in two modes: Local Shadow Parsing for offline access and Cloud-based Official SDK for authoritative data. Use Local Mode to quickly explore your Mendix project without leaving your machine, and switch to Cloud Mode when you need the most accurate representation from the Mendix Cloud.

How to install

Prerequisites are required before you begin.

1. Install Node.js v18 or higher.

2. Ensure you have a local Mendix project available.

3. Clone the project and install dependencies.

git clone https://github.com/YourUsername/mendix-local-mcp.git
cd mendix-local-mcp
npm install

4. Build the project.

npm run build

Additional setup for cloud access

If you plan to use the Official SDK in Cloud Mode, configure your Mendix Personal Access Token (PAT) and connect the server to your MCP client.

{
  "mcpServers": {
    "mendix-local-mcp": {
      "command": "node",
      "args": ["/absolute/path/to/mendix-local-mcp/build/server.js"],
      "env": {
        "MENDIX_TOKEN": "your_generated_pat_string",
        "MENDIX_USERNAME": "[email protected]"
      }
    }
  }
}

Configuration

To enable cloud access, add your MCP server configuration to your MCP client configuration (for example, a file like mcp_config.json) with the exact command and environment variables shown in the example below.

Configure the server using this exact snippet in your MCP client config.

{
  "mcpServers": {
    "mendix-local-mcp": {
      "command": "node",
      "args": ["/absolute/path/to/mendix-local-mcp/build/server.js"],
      "env": {
        "MENDIX_TOKEN": "your_generated_pat_string",
        "MENDIX_USERNAME": "[email protected]"
      }
    }
  }
}

Security and privacy

You can run the server in Local Shadow Parsing mode to keep your Mendix data offline on your machine. When you need authoritative data, you can enable the Official SDK in Cloud Mode using your Mendix PAT and username.

Troubleshooting

If you encounter issues starting the server, verify that Node.js is installed, the build output exists at the expected path, and your environment variables are correct. Ensure your Mendix project is accessible to the process and that the PAT has the required scope for model repository reads.

Project structure

The source tree includes modules for the MCP server, shadow parsing, and DTO mappings. Core files include the entry point, the server logic, the SQLite reader for the shadow mode, binary parsers for Mendix units, and DTO definitions for the official SDK.

Available tools

list_local_documents

Lists all documents in the Mendix project; can filter by module to narrow results.

get_domain_model

Extracts a simplified Domain Model by parsing Mendix project data.

inspect_local_microflow

Reads local binary definitions to reveal microflow logic.

inspect_database_schema

Debug tool to inspect the internal .mpr SQLite schema.

get_module_entities_sdk

Fetches an accurate, sanitized Domain Model representation from Mendix Cloud using the Official SDK.