Asking your AI to deploy NPL

Learn how to interact with your NOUMENA Cloud NPL applications using an MCP Server integrated editor.

Prerequisites

Before you begin, ensure you have:

• Installed an MCP enabled editor, for example VS Code or Cursor
• Installed npl-cli MCP server. See VS Code and Cursor integration

This also assumes that you have already created the following NOUMENA Cloud components:

• An NPL Project. See Creating a new NPL project
• A tenant. See Subscribe to NOUMENA Cloud
• A NOUMENA Cloud application. For a new app, see Create an application

Use MCP to build and deploy NPL

The MCP server exposes the functionality of the NPL CLI as services.

Local services that are available via MCP:

• check - syntactically verify NPL code
• test - Run all defined tests
• openapi - generate openapi specs
• puml - generate a PUML diagram
• deploy - deploy NPL sources to a local engine

Cloud services available via MCP:

• login - log into a Cloud instance
• logout - logout of a Cloud instance
• clear - Reset NPL cloud instance
• cloud deploy npl - deploy NPL sources to a remote engine
• cloud deploy frontend - deploy front end source codee

To use any of these services, simply open the command pane in your MCP editor and describe the desired action. Follow whatever prompts are given.

Examples:

Check npl code or Verify NPL Code

Deploy my application to my local engine

Deploy my app using tenant 'my-tenant' and app 'my-app'

Deploy my frontend app

Workflows

MCP services prove handy in typical NPL development workflows, documented here:

• Developing and running NPL locally

• Deploying NPL and frontends to an application running in the NOUMENA Cloud

Next steps

Now that you have built and deployed your NPL applications using MCP, explore:

• Coding in NPL with an AI agent
• Developing in Codespaces
• Using NPL applications