🌌 Hello MCP - Star Wars API Tools

Welcome to Hello MCP, a Node.js project that provides tools to interact with the Star Wars API (SWAPI). 🚀 Whether you're a Star Wars fan or a developer looking to explore the galaxy far, far away, this project has you covered! 🌠

---

About

This is a server and client implementation of the MCP protocol. It wraps the Swapi.dev API.

✨ Features

• 🔍 Character Height Comparison
  Find out if a Star Wars character is shorter or taller than a stormtrooper.

• 🌍 Planet Information
  Retrieve details about Star Wars planets, including gravity and population.

• 🌍🌍🌍 Planets Information
  Retrieve details about planets in Star Wars.

• ⚡ Fast and Easy to Use
  Built with Node.js for quick and efficient API interactions.

---

🚀 Getting Started

Follow these steps to get up and running:

1. Clone the Repository

   git clone https://github.com/your-username/hello-mcp.git
   cd hello-mcp

2. Install Dependencies
   Make sure you have Node.js installed, then run:

   npm install

3. Run the Project

   Start the server with:

   npm run start:server

   Start the client with:

   npm run start:client

   The client runs code like below:

   const query = "Is Luke shorter than a stormtrooper?";
   const result = await client.processQuery(query);
   console.log("Final result: \n ", result);

   The preceding code would trigger the tool defined in /src/tools/shorter.ts.

🛠️ Tools Overview

If starting the server with npm run start:server, the client will contain the tools:

• Character Tool
  Endpoint: /tools/call_swapi_character
  Input: { "name": "Luke Skywalker" }
  Output: "Luke Skywalker is shorter than a stormtrooper."

• Planet Tool
  Endpoint: /tools/call_swapi_planet
  Input: { "name": "Tatooine" }
  Output: "Tatooine has gravity 1 standard with a population of 200000."

If starting the server with npm run start:chuck, the client will contain the tools:

• Chuck Norris Random Joke Tool
  Endpoint: /tools/random-joke Input: {} Output: "Chuck Norris joke: Chuck Norris can divide by zero."

• Chuck Norris Random Joke by Category Tool
  Endpoint: /tools/random-joke-by-category Input: {"category": "Animal"} Output: "Chuck Norris joke: Chuck Norris can divide by zero."

How it works

The client negotiates with the server for its available tools. It will figure out whether to call both tool and getting a generic LLM response. The end result will be a combination of both if available.

Adding a tool

A tool is created by calling the server.tool(<toolname>, <description>, { /* implementation */ })

• Locate one of the tools in src/tools and create a new one like so:

  o src/tools/exampleTool.ts o Add tool to register it in import section in src/tools/server.ts o Write a query in src/client.ts that semantically matches the description you've created for the tool, for example: o query: "add 1 to 2" o tool description: "adding two numbers"

🤝 Contributing

Contributions are welcome! Feel free to open issues or submit pull requests to improve the project.

---

📜 License

This project is licensed under the MIT License.

---

May the Force be with you! 🌟

Adding to Visual Studio Code - Agent mode

For Codespaces .vscode/mcp.json

{
    "inputs": [],
    "servers": {
       "hello-mcp": {
           "command": "sh",
           "args": [
               "-c",
               "cd /workspaces/hello-mcp && ./node_modules/.bin/tsc && node ./build/server.js"
           ]
       }
    }
}

Thank you

• Than you Wassim Chegham for the original code. Also please check out his repo at https://github.com/manekinekko/openai-mcp-example