# My First MCP Server: Community Toolkit 🛠️

A beginner-friendly [Model Context Protocol](https://modelcontextprotocol.io/) server built by a DevRel person learning MCP from scratch. If I can build this, you can too.

## What it does

Gives Claude (or any MCP client) three tools every community builder needs:

| Tool | What it does |
|------|-------------|
| `days_until_event` | Countdown to your next meetup |
| `meetup_announcement` | Generates a ready-to-post Markdown announcement |
| `thank_you_note` | Drafts contributor appreciation notes |

## Quick start (5 minutes)

**1. Install**

```bash
pip install "mcp[cli]"
```

**2. Test it with the MCP Inspector**

```bash
mcp dev server.py
```

This opens a browser UI where you can call each tool and see the responses — no AI client needed yet.

**3. Connect to Claude Desktop**

Add to your `claude_desktop_config.json`:

```json
{
  "mcpServers": {
    "community-toolkit": {
      "command": "python",
      "args": ["/full/path/to/server.py"]
    }
  }
}
```

Restart Claude Desktop, then try: *"How many days until my meetup on 2026-09-15?"*

## What I learned building this

- MCP is genuinely "USB-C for AI" — I wrote plain Python functions and the protocol handled everything else.
- The `@mcp.tool()` decorator + a good docstring is all the AI needs to know when to use your tool.
- The MCP Inspector (`mcp dev`) is the fastest feedback loop for beginners.

## Next steps

- Add a tool that pulls real data (e.g., Meetup/Luma API)
- Try resources and prompts, not just tools

Built as part of my [AAIF Ambassador](https://aaif.io/ambassadors/) journey. Blog post: *coming soon*.