convez.ai
MCP Server
garland3

mcp-server-template

garland3/mcp-server-template

Ver no GitHub
InicialPluginScore · de 0 a 100

Instalação

$npx convez-ai garland3/mcp-server-template

1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.

Métricas

Estrelas
0
Forks
0
Acompanhando
0
Saúde
14%
Último commit
há 5 meses
Versão
—
Issues abertas
0
Contribuidores
2

README

A tradução para o português deste README ainda está na fila.

my-module

Template repo. This is the team's starting point for new MCP servers. Clone it, rename my_module, and fill in the blanks below. It already satisfies all six MCP repo requirements: layout, pinned Python, this README, an about prompt, a passing test, and the .gitignore.

1. What does this do, and when should I use it?

This is a template MCP (Model Context Protocol) server built on FastMCP. It exposes one example tool, hello_world, which returns a greeting, plus an about prompt that tells the connecting LLM what the server is for. Use it as the skeleton for any new MCP — replace the example tool with your real computation in core.py.

2. How do I install it from nothing?

Assumes zero setup. Install uv if you don't have it, then:

# 1. Install uv (skip if already installed)
curl -LsSf https://astral.sh/uv/install.sh | sh

# 2. Clone the repo
git clone https://github.com/garland3/mcp-server-template.git
cd mcp-server-template

# 3. Install dependencies (creates a .venv automatically)
uv sync --extra dev

# 4. Confirm it works
uv run pytest          # tests should pass
uv run my-module       # starts the MCP server (Ctrl-C to stop)

3. How do I connect it?

Point your MCP client at the example config in mcp.json. Edit the --directory path to the absolute path of this repo on your machine:

{
  "mcpServers": {
    "my-module": {
      "command": "uv",
      "args": ["run", "--directory", "/absolute/path/to/this/repo", "my-module"]
    }
  }
}

Drop that into your client's MCP settings (e.g. Claude Desktop's claude_desktop_config.json) and restart the client. The server will offer the hello_world tool and the about prompt.

4. Screenshot

Replace the image below with a screenshot of your MCP working in a client.

Screenshot of the MCP server working


Project layout

root
├── pyproject.toml        # dependencies + install (Python >=3.11)
├── mcp.json              # example connection config
├── README.md             # this file
├── .gitignore            # keeps credentials/large data out of the repo
├── src/
│   └── my_module/
│       ├── main.py       # MCP server wiring (FastMCP)
│       ├── core.py       # the actual computation
│       └── About.md      # text served as the `about` MCP prompt
└── tests/
    └── test_core.py      # at least one test that runs

Keep computation in core.py and MCP wiring in main.py. The about prompt is loaded from About.md, so you can update what the LLM is told about the server without touching any code.

Ver o original no GitHub

Relacionados