mermaid-mcp
A basic MCP server example with a hello tool, info resource, and assistant prompt, built with TypeScript and Node.js using the official MCP SDK.
README
MCP Server Básico
Servidor Model Context Protocol (MCP) de exemplo, escrito em TypeScript e Node.js, usando o SDK oficial @modelcontextprotocol/sdk.
📖 Nunca usou MCP? Siga o Guia de Uso passo a passo — do zero até usar no Claude Desktop, sem precisar saber programar.
Capacidades
| Tipo | Nome | Descrição |
|---|---|---|
| Tool | hello |
Recebe name e retorna Olá, <name>! |
| Resource | info |
resource://info — texto explicando o servidor |
| Prompt | assistant |
Recebe topic e gera um prompt de sistema sobre esse tema |
Estrutura do projeto
mermaid MCP/
├── src/
│ ├── index.ts # Ponto de entrada: cria o servidor e conecta via stdio
│ ├── tools/
│ │ └── hello.ts # Tool "hello"
│ ├── resources/
│ │ └── info.ts # Resource "info" (resource://info)
│ └── prompts/
│ └── assistant.ts # Prompt "assistant"
├── package.json
├── tsconfig.json
└── README.md
Arquitetura
graph TD
Claude["Cliente Claude<br/>(Claude Desktop)"]
Server["MCP Server<br/>(index.ts — stdio / JSON-RPC 2.0)"]
Tools["Tools<br/>(src/tools/)"]
Resources["Resources<br/>(src/resources/)"]
Prompts["Prompts<br/>(src/prompts/)"]
Hello["hello<br/>name → 'Olá, name!'"]
Info["info<br/>resource://info"]
Assistant["assistant<br/>topic → prompt de sistema"]
Claude -- "tools/call" --> Server
Claude -- "resources/read" --> Server
Claude -- "prompts/get" --> Server
Server --> Tools
Server --> Resources
Server --> Prompts
Tools --> Hello
Resources --> Info
Prompts --> Assistant
Hello -. "resposta (content)" .-> Server
Info -. "resposta (contents)" .-> Server
Assistant -. "resposta (messages)" .-> Server
Server -. "resultado JSON-RPC" .-> Claude
Fluxo de uma chamada de Tool
sequenceDiagram
participant Claude as Cliente Claude
participant Server as MCP Server
participant Tool as Tool "hello"
Claude->>Server: initialize
Server-->>Claude: capabilities (tools, resources, prompts)
Claude->>Server: notifications/initialized
Claude->>Server: tools/call { name: "hello", arguments: { name: "Maria" } }
Server->>Tool: valida input (Zod) e executa handler
Tool-->>Server: { content: [{ type: "text", text: "Olá, Maria!" }] }
Server-->>Claude: resposta JSON-RPC com o resultado
Requisitos
- Node.js 18 ou superior
Instalação
npm install
Desenvolvimento
Executa direto do TypeScript, sem compilar (via tsx):
npm run dev
Build e execução
npm run build # compila TypeScript para ./build
npm start # executa node build/index.js
O servidor se comunica por stdio (stdin/stdout com JSON-RPC 2.0). Ele não abre porta HTTP — quem o inicia é o cliente MCP (ex.: Claude Desktop).
Testando manualmente
Com o MCP Inspector:
npx @modelcontextprotocol/inspector node build/index.js
Conectando ao Claude Desktop
- Compile o projeto:
npm run build - Abra o arquivo de configuração do Claude Desktop:
- Windows:
%APPDATA%\Claude\claude_desktop_config.json - macOS:
~/Library/Application Support/Claude/claude_desktop_config.json
- Windows:
- Adicione o servidor (ajuste o caminho absoluto):
{
"mcpServers": {
"mcp-server-basico": {
"command": "node",
"args": [
"C:\\CAMINHO\\ATE\\O\\PROJETO\\mermaid-mcp\\build\\index.js"
]
}
}
}
- Reinicie o Claude Desktop. O servidor aparecerá no menu de ferramentas (ícone 🔌 / "Search and tools").
Depois disso você pode pedir, por exemplo: "Use a tool hello com o nome Maria".
Adicionando novas Tools
- Crie um arquivo em
src/tools/, exportando uma funçãoregisterXxxTool(server). - Dentro dela, chame
server.registerTool(nome, config, handler). - Importe e chame a função em
src/index.ts. - Rode
npm run builde reinicie o cliente MCP.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.