- Explore MCP Servers
- ai-agent-with-mcp
Ai Agent With Mcp
What is Ai Agent With Mcp
ai-agent-with-mcp is a project that implements an AI agent utilizing the Model Context Protocol (MCP) to interact with external resources.
Use cases
Use cases include creating chatbots that can greet users, managing user information in applications, and integrating with external APIs to enhance user interactions.
How to use
To use ai-agent-with-mcp, ensure you have Cursor installed and npm configured. You can install it via the terminal using the command: npx -y @smithery/cli@latest install @wonderwhy-er/desktop.
Key features
Key features include user resource management through the api://users endpoint, which lists registered users with details such as ID, name, phone number, and thread ID, and a simple greeting message resource available at hello://world.
Where to use
ai-agent-with-mcp can be used in various fields such as customer service, chatbots, and any application requiring interaction with user data and external resources.
Clients Supporting MCP
The following are the main client software that supports the Model Context Protocol. Click the link to visit the official website for more information.
Overview
What is Ai Agent With Mcp
ai-agent-with-mcp is a project that implements an AI agent utilizing the Model Context Protocol (MCP) to interact with external resources.
Use cases
Use cases include creating chatbots that can greet users, managing user information in applications, and integrating with external APIs to enhance user interactions.
How to use
To use ai-agent-with-mcp, ensure you have Cursor installed and npm configured. You can install it via the terminal using the command: npx -y @smithery/cli@latest install @wonderwhy-er/desktop.
Key features
Key features include user resource management through the api://users endpoint, which lists registered users with details such as ID, name, phone number, and thread ID, and a simple greeting message resource available at hello://world.
Where to use
ai-agent-with-mcp can be used in various fields such as customer service, chatbots, and any application requiring interaction with user data and external resources.
Clients Supporting MCP
The following are the main client software that supports the Model Context Protocol. Click the link to visit the official website for more information.
Content
AI Agent with MCP
Este projeto implementa um agente de IA que utiliza o Model Context Protocol (MCP) para interagir com recursos externos.
Recursos Disponíveis
O servidor MCP fornece os seguintes recursos:
Recursos de Usuários
- api://users: Lista todos os usuários cadastrados no sistema
- Retorna informações como ID, nome, número de telefone e thread ID
- Formato: JSON
- Exemplo de resposta:
{ "content": [ { "id": 1, "phoneNumber": "553496341404", "name": "Moisés Paschoalick", "threadId": "thread_bZEPVYVBvHxY9Ok6WqR63M2D" }, { "id": 2, "phoneNumber": "553496338888", "name": "José Silva", "threadId": "1thread_t8SFjKio6yN9pppqypilwGoR__" }, { "id": 3, "phoneNumber": "553496338889", "name": "Maria Antonia", "threadId": "1thread_t8SFjKio6yN9pppqypilwGoR__" } ], "pageable": { "pageNumber": 0, "pageSize": 12, "sort": { "empty": false, "sorted": true, "unsorted": false }, "offset": 0, "paged": true, "unpaged": false }, "last": true, "totalPages": 1, "totalElements": 3, "first": true, "size": 12, "number": 0, "sort": { "empty": false, "sorted": true, "unsorted": false }, "numberOfElements": 3, "empty": false }
Recursos de Mensagens
- hello://world: Retorna uma mensagem de saudação simples
- Formato: texto plano (text/plain)
- Exemplo de resposta: “Hello, World! This is my first MCP resource.”
Instalação para Cursor
Primeiro, certifique-se de que você tem o Cursor instalado e o npm configurado em seu sistema.
Opção 1: Instalação via Terminal
Execute o seguinte comando no terminal:
npx -y @smithery/cli@latest install @wonderwhy-er/desktop-commander --client cursor --key dda23bec-caa6-4487-a1e9-eb74e22e33eb
Reinicie o Cursor se estiver em execução.
Opção 2: Instalação Manual
Adicione a entrada apropriada ao seu arquivo mcp.json:
Para Mac/Linux:
No Linux: ~/.config/cursor.json
{
"mcpServers": {
"desktop-commander": {
"command": "npx",
"args": [
"-y",
"@smithery/cli@latest",
"run",
"@wonderwhy-er/desktop-commander",
"--key",
"dda23bec-caa6-4487-a1e9-eb74e22e33eb"
]
}
}
}
Para Windows:
{
"mcpServers": {
"desktop-commander": {
"command": "cmd",
"args": [
"/c",
"npx",
"-y",
"@smithery/cli@latest",
"run",
"@wonderwhy-er/desktop-commander",
"--key",
"dda23bec-caa6-4487-a1e9-eb74e22e33eb"
]
}
}
}
Reinicie o Cursor se estiver em execução.
Opção 3: Clonar localmente
Clone e construa:
git clone https://github.com/moises-paschoalick/ai-agent-with-mcp.git
cd ai-agent-with-mcp
npm run setup
Reinicie o Cursor se estiver em execução.
O comando de configuração irá:
- Instalar dependências
- Construir o servidor
- Configurar o Cursor
- Adicionar servidores MCP à configuração do Cursor, se necessário
Como Usar
Cliente MCP
O projeto inclui um cliente MCP que pode ser usado para acessar os recursos disponíveis:
import { Client } from "@modelcontextprotocol/sdk/client/index.js";
import { StdioClientTransport } from "@modelcontextprotocol/sdk/client/stdio.js";
async function main() {
const client = new Client({
name: "hello-mcp-client",
version: "1.0.0"
});
const transport = new StdioClientTransport({
command: "node",
args: ["build/index.js"]
});
try {
await client.connect(transport);
// Listar recursos disponíveis
const resources = await client.listResources();
console.log("Recursos disponíveis:", resources);
// Ler o recurso de usuários
const content = await client.readResource({ uri: "api://users" });
console.log("\nLista de Usuários:", content);
} catch (error) {
console.error("Erro ao executar o cliente:", error);
}
}
main().catch(console.error);
API REST Direta
Você também pode acessar os recursos diretamente via API REST:
# Listar todos os usuários
curl http://3.238.149.189:8080/users
Instalação e Execução
# Instalar dependências
npm install
# Compilar o projeto
npm run build
# Executar o servidor
npm start
# Executar em modo de desenvolvimento
npm run dev
Requisitos
- Node.js (versão compatível com AbortController)
- TypeScript
- @modelcontextprotocol/sdk
Dev Tools Supporting MCP
The following are the main code editors that support the Model Context Protocol. Click the link to visit the official website for more information.










