LabVIEW -Toolkit

This LabVIEW MCP Server Toolkit exposes a local HTTP MCP server that describes tools, prompts and resources you can control from an MCP client (for example Claude Desktop). The server opens on the current machine and communicates using JSON-RPC 2.0 over HTTP POST. Clients can discover the available LabVIEW-related tools (VI calls, automation actions, data retrievals) and then request the LLM to execute specific VIs or workflows in LabVIEW. Tools are defined with inputs, outputs, a name and a description so the AI can decide how to invoke them, while prompts and resources provide configurable templates and payloads the AI can reuse during a chat session. To connect from a client, configure an MCP server entry similar to the example, pointing the client at http://127.0.0.1:36987/mcp/server (or the port you configure). Once connected, the AI can propose tool invocations like running a VI with particular parameters or retrieving status information from LabVIEW.

Prerequisites:

• Node.js (with npm) installed on the machine that runs LabVIEW
• Access to LabVIEW environment where VIs can be executed via MCP tools

Installation steps:

1. Install Node.js from https://nodejs.org if you don’t have it already.
2. Install or install locally the LabVIEW MCP Toolkit project (clone/download from GitHub).
3. Open a terminal in the project directory and install dependencies (if a package.json is provided):

npm install

4. Start the MCP server. The README example shows using npx with mcp-remote to expose http://127.0.0.1:36987/mcp/server. You can start it with:

npx mcp-remote http://127.0.0.1:36987/mcp/server

5. If you run behind a firewall or want to change the port, adjust the server configuration accordingly and restart.
6. In your MCP client (e.g., Claude Desktop), add a new MCP server entry with the following:

{
  "mcpServers": {
    "vi-scripting-server": {
      "command": "npx",
      "args": [
        "mcp-remote",
        "http://127.0.0.1:36987/mcp/server"
      ],
      "env": {}
    }
  }
}

Notes:

• Ensure the LabVIEW environment is accessible to the process running the MCP server.
• The port in the URL should match what you configure when starting the server.
• For first-time setup, you may need to restart the MCP client after starting the server.

Tips and common issues:

• If the MCP client cannot connect, verify that the server is listening on the expected port and that there are no firewall blocks.
• If tools/prompts/resources do not show up, ensure the server has loaded the LabVIEW VI bindings and that the HTTP endpoint is reachable from the client machine.
• You may want to run the server with verbose logging to troubleshoot: check for JSON-RPC request handling and any LabVIEW VI invocation errors.
• The npm package used to run the server in the example is mcp-remote; this is the recommended entry point for starting the MCP server in a Node.js environment.
• When editing prompts or tools, keep the input/output schemas aligned with MCP 2.0 expectations to ensure the client can render UI controls correctly.