Skip to main content
Glama
stevennevins

MCP Server Template

by stevennevins

Plantilla de servidor MCP

Una plantilla para crear servidores de Protocolo de Contexto de Modelo (MCP) en TypeScript. Esta plantilla proporciona una base sólida para construir servidores compatibles con MCP con las herramientas adecuadas, seguridad de tipos y las mejores prácticas.

Características

  • 🚀 Compatibilidad total con TypeScript

  • 🏗️ Inyección de dependencia basada en contenedores

  • 📦 Arquitectura basada en servicios con interfaz DataProcessor

  • 🛠️ Ejemplo de implementación de la herramienta con pruebas

  • Marco de pruebas Vitest

  • 📝 Definiciones de tipos

  • 🔌 Integración del SDK de MCP

Related MCP server: MCP Server Template

Empezando

Desarrollo

  1. Instalar dependencias:

    npm install
  2. Inicie el servidor de desarrollo con recarga en caliente:

    npm run dev
  3. Construir el proyecto:

    npm run build
  4. Ejecutar pruebas:

    npm test
  5. Inicie el servidor de producción:

    npm start

Estructura del proyecto

src/
├── index.ts          # Entry point
├── interfaces/       # Interface definitions
│   └── tool.ts      # DataProcessor interface
└── tools/           # Tool implementations
    └── example.ts   # Example tool

Creación de herramientas

  1. Exporte su herramienta y controladores siguiendo el ejemplo en src/tools/example.ts :

    // In your-tool.ts
    export const YOUR_TOOLS = [
      {
        name: "your-tool-name",
        description: "Your tool description",
        parameters: {
          // Your tool parameters schema
        },
      },
    ];
    
    export const YOUR_HANDLERS = {
      "your-tool-name": async (request) => {
        // Your tool handler implementation
        return {
          toolResult: {
            content: [{ type: "text", text: "Result" }],
          },
        };
      },
    };
  2. Registre su herramienta en las constantes ALL_TOOLS y ALL_HANDLERS en src/index.ts :

    // In src/index.ts
    import { YOUR_TOOLS, YOUR_HANDLERS } from "./tools/your-tool.js";
    
    // Combine all tools
    const ALL_TOOLS = [...EXAMPLE_TOOLS, ...YOUR_TOOLS];
    const ALL_HANDLERS = { ...EXAMPLE_HANDLERS, ...YOUR_HANDLERS };

El servidor automáticamente:

  • Enumere su herramienta en las herramientas disponibles

  • Manejar la validación de entrada

  • Procesar solicitudes a su herramienta

  • Formatear respuestas según el protocolo MCP

Pruebas

La plantilla incluye un TestClient integrado para pruebas locales y el MCP Inspector para depuración visual.

Uso de TestClient

TestClient proporciona una forma sencilla de probar sus herramientas:

import { TestClient } from "./utils/TestClient";

describe("YourTool", () => {
  const client = new TestClient();

  it("should process data correctly", async () => {
    await client.assertToolCall(
      "your-tool-name",
      { input: "test" },
      (result) => {
        expect(result.toolResult.content).toBeDefined();
      }
    );
  });
});

Uso del inspector MCP

La plantilla incluye el Inspector MCP para la depuración visual de sus herramientas:

  1. Iniciar el inspector:

    npx @modelcontextprotocol/inspector node dist/index.js
  2. Abra la interfaz de usuario del inspector en http://localhost:5173

El inspector proporciona:

  • Interfaz visual para herramientas de prueba

  • Monitoreo de solicitudes y respuestas en tiempo real

  • Inspección de metadatos de la herramienta

  • Entorno de pruebas interactivo

Pruebas locales con cursor

Para probar su servidor MCP localmente con Cursor:

  1. Construya y vincule el paquete:

    npm run build
    npm run link
  2. Verificar que el binario funciona:

    npx example-mcp-tool
  3. Añade el servidor al Cursor:

    • Abrir la configuración del cursor

    • Vaya a la pestaña Características

    • Desplácese hacia abajo hasta la sección Servidores MCP

    • Haga clic en "Agregar servidor"

    • Seleccione el tipo "Comando"

    • Asígnele un nombre (por ejemplo, "Herramienta de ejemplo local")

    • Ingrese el comando: npx example-mcp-tool

    • Haga clic en Confirmar

  4. Verifique que el servidor se inicie correctamente en Cursor verificando que la sección Servidores MCP muestre que su servidor está en ejecución.

Nota: Si realiza cambios en su código, recuerde reconstruirlo y volver a vincularlo:

npm run build
npm run link

Cuando hayas terminado de probar, puedes desvincular el paquete:

npm run unlink

Esto eliminará el enlace simbólico global creado durante el desarrollo.

Available Tools

1 tool
example-toolC

An example tool that processes input data

ParametersJSON Schema
NameRequiredDescriptionDefault
inputYesInput string to process

TDQS

C2.6/5.0
Behavior2/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are provided, so the description carries the full burden of behavioral disclosure. It states the tool 'processes input data', which implies some action but doesn't reveal behavioral traits like whether it's read-only, destructive, requires authentication, has side effects, or rate limits. This leaves significant gaps in understanding how the tool behaves.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is a single, efficient sentence with no wasted words, making it appropriately concise. However, it's front-loaded with the basic purpose but lacks structure or additional details that could enhance clarity, keeping it simple but under-specified.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness2/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Given the tool's simplicity (1 parameter, no output schema, no annotations), the description is incomplete. It doesn't explain what 'processes' means, the expected output, or behavioral context, leaving the agent with insufficient information to use the tool effectively despite the straightforward schema.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters3/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The description adds no specific meaning about the 'input' parameter beyond what the schema provides, which has 100% coverage and describes it as 'Input string to process'. With high schema coverage, the baseline is 3, as the schema adequately documents the parameter, and the description doesn't compensate or add further semantics.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose3/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description states the tool 'processes input data', which provides a basic purpose but is vague about what 'processes' entails. It doesn't specify the type of processing or outcome, and with no sibling tools, differentiation isn't needed. This is a minimal viable description that communicates a general function without specifics.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines2/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description offers no guidance on when to use this tool, such as context, prerequisites, or alternatives. With no sibling tools, it doesn't need to distinguish from others, but it lacks any usage instructions or scenarios, leaving the agent without direction on appropriate application.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

  1. 1 tool update
    • First observedexample-tool

TDQS

C2.8/5.0

Scored across 1 tool

Disambiguation5/5

With only one tool, there is no possibility of confusion or overlap between tools, as there are no other tools to compare it against. The single tool's purpose is clearly defined, eliminating any ambiguity in tool selection.

Naming Consistency5/5

Since there is only one tool, naming consistency is inherently perfect—there are no other tools to create inconsistency. The tool name 'example-tool' follows a simple, clear pattern without any conflicting conventions to evaluate.

Tool Count2/5

A single tool is generally too few for a server's purpose, as it limits functionality and suggests an incomplete or trivial implementation. For a server named 'MCP Server Template', one tool feels insufficient to demonstrate a coherent set of capabilities, making it borderline inadequate.

Completeness1/5

The server's purpose is unclear from the name 'MCP Server Template', but with only one generic tool ('example-tool'), there are significant gaps in coverage. It lacks any CRUD operations, lifecycle management, or domain-specific functions, making it severely incomplete for any practical application.

Maintenance

ActivityInactive
ResponsivenessNo issues

Related MCP Connectors

Related MCP Servers