Building with AI

Kiro Power

The Nx Plugin for AWS ships with a Kiro Power that bundles the MCP server along with documentation and workflow guides, giving Kiro rich context about the available generators and how to use them. This is the recommended way to get started with Kiro.

Installing the Power

Open the Kiro Powers panel from the sidebar
Click the + button to add a custom power

Paste the following GitHub URL:

https://github.com/awslabs/nx-plugin-for-aws/tree/main/powers/nx-plugin-for-aws

Click install

Once installed, Kiro automatically has access to the aws-nx-mcp MCP server and its tools. No additional MCP configuration is needed.

What the Power Provides

The power gives Kiro access to the following MCP tools:

create_workspace_command — Generate the command to create a new Nx workspace pre-configured with the AWS plugin
list_generators — Discover all available generators and their parameters
generator_guide — Get detailed documentation for a specific generator, including parameters, file structure, and best practices
general_guidance — Get general guidance on using the plugin

It also includes steering documentation that helps Kiro understand the full set of available generators, common workflows, and best practices.

Example Prompts

Once the power is installed, you can start building by chatting with Kiro:

“Create a new Nx workspace called my-app using pnpm” — Kiro will generate and run the appropriate workspace creation command
“Scaffold a React website with a tRPC API backend and CDK infrastructure” — Kiro will scaffold each project and connect them together
“What generators are available in the Nx Plugin for AWS?” — Kiro will list all generators with descriptions and example usage
“Add a Lambda function to my existing TypeScript project” — Kiro will find the right generator and walk through the scaffolding process
“Build my project” - Kiro will execute the nx build command

MCP Server Configuration

The Nx Plugin for AWS includes an MCP Server which enables AI assistants to work with the Nx Plugin for AWS. By utilizing the MCP Server, you can accelerate your development workflow with AI, while benefitting from your projects being scaffolded in a deterministic fashion, spending less time and less of your context window on setting up the main components.

Use the MCP Server with your AI Assistant of choice, such as Kiro, Amazon Q Developer, Cline, Claude Code or Cursor.

For other AI Assistants, you can configure the MCP server manually. Most AI Assistants will have a JSON file for MCP server configuration. In Amazon Q Developer, this is located in ~/.aws/amazonq/mcp.json. Add an entry for the Nx Plugin for AWS MCP Server:

{
  "mcpServers": {
    "nx-plugin-for-aws": {
      "command": "npx",
      "args": ["-y", "-p", "@aws/nx-plugin", "aws-nx-mcp"]
    }
  }
}

Please refer to the following documentation for configuring MCP with specific AI Assistants:

Start Vibe Coding

Ask your AI Assistant to build something using the nx-plugin-for-aws! Generally AI Assistants will start by creating a workspace, scaffolding with the applicable generators, and then filling in the business logic.

Build your own MCP Server

Check out the ts#mcp-server generator guide for details about building your own MCP Server.