home / mcp / new mcp builder mcp server

New MCP Builder MCP Server

Starter MCP server for building tools, resources, and prompts with multi-client integration.

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "boosclues-new-mcp-builder": {
      "url": "https://example-mcp-endpoint.invalid",
      "headers": {
        "API_KEY": "your-api-key",
        "ENV_VARS": "Included from .env.local",
        "DATABASE_URL": "postgres://user:pass@host/db"
      }
    }
  }
}

You can quickly spin up an MCP server boilerplate to build tools, resources, and prompts that integrate with Claude Desktop, Cursor, Claude Code, Gemini, and other MCP-compatible AI assistants. This server serves as a clean foundation you can customize to expose your own endpoints and workflows.

How to use

Use this MCP server to create callable tools and resources that your AI assistants can invoke. You can run it directly with an MCP client via npx, or run it locally for development and testing. The server is designed to publish updates to MCP clients automatically and to work with a local development flow that includes building and starting the server.

How to install

Prerequisites you need on your machine before starting:

Node.js 14+ (with npm) or a compatible Node.js runtime
pnpm (recommended) or npm for package installation

Option 1 — Use the published package (recommended):

npx @r-mcp/boilerplate@latest

Option 2 — Customize and develop locally:

# Clone the boilerplate
git clone <your-repo-url>
cd mcp-server-boilerplate

# Install dependencies
pnpm install

# Build the project
pnpm run build

# Start the server
pnpm start

Installation scripts are provided to deploy to multiple MCP clients and to set up local development. You can install to all clients or to specific ones as shown below.

# Install to all MCP clients (Claude Desktop, Cursor, Claude Code, Gemini, MCP)
pnpm run install-server

# Install to specific clients
pnpm run install-desktop       # Claude Desktop
pnpm run install-cursor        # Cursor IDE
pnpm run install-code          # Claude Code CLI
pnpm run install-code-library  # Claude Code Library (~/.claude/mcp-library/)
pnpm run install-mcp           # Local .mcp.json for development

# You can also combine multiple targets
node scripts/update-config.js cursor code desktop

Additional steps to run and test

After you install, start the server and connect with your MCP client. The production configuration uses npx to fetch the latest boilerplate package, while a local development setup runs the compiled local server via node with an absolute path to dist/index.js.

Customizing your server

This boilerplate is designed to be extended with new tools, resources, and prompts. You can add your own tool definitions, dynamic resources, and reusable prompts to tailor the MCP server to your use cases.

Environment and publishing notes

Create a .env.local to supply environment-specific configuration. These variables are included in MCP server configurations during installation and can be used by your tools and resources.

When you are ready to publish updates, you can build, commit, and publish with a single command. This workflow increments the patch version, builds the project, and pushes the update to the MCP ecosystem so clients auto-update.

Project structure at a glance

Key parts you’ll work with include the main server entry, build scripts, and a distributed output ready for MCP client integration.

Development workflow

Local development keeps changes fast and testable. Build after changes, run the local server, and verify behavior with your MCP client. When ready, publish updates so clients receive the latest version automatically.

Environment variables

Create a .env.local file for environment-specific configuration, such as API keys or database connections. These variables are automatically included in MCP server configurations during installation.

Next steps

1) Create your own server name and customize tools in the source. 2) Add tools, resources, and prompts to expand capabilities. 3) Configure environment variables in .env.local. 4) Build and publish your server so MCP clients auto-update. 5) Install to clients with the provided installation commands.

Available tools

hello-world

A simple tool that returns a greeting and demonstrates parameter validation using a Zod schema.

get-mcp-docs

Retrieves MCP documentation or usage guidelines and formats them for display by the AI assistant.