Multi Context Protocol (MCP) Server
LocalAbout Multi Context Protocol (MCP) Server
A travel-planning MCP server for questions involving flights, hotels, and timing. It provides flight searches through Booking.com's API, hotel searches, and the current date for AI clients.
Key features
- 01Search available flights
- 02Search hotels and accommodations
- 03Provide the current date
Use cases
- 01Find a flight arriving before a set time
- 02Find hotels within walking distance of a landmark
MCP support
Exposes flight search, hotel search, and current-date functions to MCP clients. The README describes use with Claude Desktop or a custom client.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
npx -y travel-mcp-serverBOOKING_COM_API_KEYSecretGOOGLE_MAPS_API_KEYSecretLocal servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.
Tools
Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.
Details
From the project README✈️ Multi Context Protocol (MCP) Server
🌟 Overview
This is a generalized travel mcp server that allows you to ask complex travel planning questions. Try the following
- I need to be at the Google office tomorrow by 4pm. What is the latest flight I need to take?
- I would like to stay at a reasonably priced hotel 15 minutes walk from the Eiffel tower, recommend some options.
🔧 Supported Functions
🛫 search-flights
Searches for available flights between two airports via Booking.com's API.
📅 today
Provides the current date to the LLM, ensuring it has up-to-date temporal context.
🏨 search-hotels
Search for available hotels and accommodations.
🔮 Features to Come
🚗 search-car-rentals
Find car rental options at your destination.
⭐ hotel-reviews
Access reviews for hotels and accommodations.
🚕 search-taxis
Find taxi and transfer services at your destination.
📦 Installation
This MCP server can be used with either Claude Desktop or your own client.
Pre-requisites
- Go to RapidAPI, and look for the booking.com API. Issue yourself a key
- Go to google cloud and issue yourself a Google Maps API key.
Use with Claude Desktop
In order to work with Claude Desktop, use the below claude_desktop_config.json
{
"mcpServers": {
"travel": {
"command": "npx",
"args": ["travel-mcp-server"],
"env": {
"BOOKING_COM_API_KEY": "<YOUR_BOOKING_DOT_COM_API_KEY>"
}
},
"google-maps": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"-e",
"GOOGLE_MAPS_API_KEY",
"mcp/google-maps"
],
"env": {
"GOOGLE_MAPS_API_KEY": "<YOUR_GOOGLE_MAPS_API_KEY>"
}
}
}
}
You will need to go get a google maps API key in order to use this.
Programmatic Use
If you want to use this programmatically, here is a code snippet.
class TravelClient {
private flightsClient: Client;
private mapsClient: Client;
private toolDefinitions: any[] = [];
constructor() {
this.flightsClient = new Client(
{ name: "FlightsApp", version: "1.0.0" },
{ capabilities: { tools: {}, resources: {}, prompts: {} } }
);
this.mapsClient = new Client(
{ name: "MapsApp", version: "1.0.0" },
{ capabilities: { tools: {}, resources: {}, prompts: {} } }
);
}
async initialize() {
console.log("Initializing MCP clients...");
try {
const flightsTransport = new StdioClientTransport({
command: "node",
args: ["../../node_modules/travel-mcp-server/build/index.js"],
});
console.log("GOOGLE MAPS API KEY", process.env.GOOGLE_MAPS_API_KEY);
const mapsTransport = new StdioClientTransport({
command: process.execPath,
args: [
"../../node_modules/@modelcontextprotocol/server-google-maps/dist/index.js",
],
env: {
GOOGLE_MAPS_API_KEY: process.env.GOOGLE_MAPS_API_KEY || "",
},
});
await Promise.all([
this.flightsClient.connect(flightsTransport),
this.mapsClient.connect(mapsTransport),
]);
// Discover tools from both servers
console.log("Discovering tools...");
const [flightsTools, mapsTools] = await Promise.all([
this.flightsClient.listTools(),
this.mapsClient.listTools(),
]);
// Combine tools from both servers
this.toolDefinitions = [
...flightsTools.tools.map((tool) => ({
name: tool.name,
description: tool.description || `Tool: ${tool.name}`,
input_schema: tool.inputSchema,
})),
...mapsTools.tools.map((tool) => ({
name: tool.name,
description: tool.description || `Tool: ${tool.name}`,
input_schema: tool.inputSchema,
})),
];
console.log(`Discovered ${this.toolDefinitions.length} tools`);
return this;
} catch (error) {
console.error("Error initializing MCP clients:", error);
throw error;
}
}
async callTool(toolName: string, parameters: any) {
const tool = this.toolDefinitions.find((t) => t.name === toolName);
if (!tool) {
throw new Error(`Tool ${toolName} not found`);
}
// Determine which client to use based on the tool name
const client = toolName.includes("maps")
? this.mapsClient
: this.flightsClient;
const toolRequest = {
name: toolName,
arguments: parameters,
};
const result = await client.callTool(toolRequest);
return result;
}
async getToolDefinitions() {
return this.toolDefinitions;
}
}
export default TravelClient;
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.