About Hotel Booking AI MCP
Hotel Booking AI MCP is a package defining the public MCP contract and companion Skill for consumer hotel booking workflows. It supports public discovery and availability checks, while booking and order operations require a user key.
Key features
- 01Hotel search and details
- 02Live single and batch rates
- 03Availability checks
- 04Booking and order management
- 05Companion Skill for agent workflows
Use cases
- 01Find hotels and inspect room rates
- 02Check availability across multiple hotels
- 03Book a hotel and manage its order
MCP support
It documents a public Streamable HTTP MCP endpoint at `https://api.tourmind.com/mcp/toc`. Search, rates, and availability need no user key; booking, order lookup, cancellation, and payment use a key supplied by the companion Skill.
Configuration
Remote servers need no local install. Pick your client, then copy the config or install in one click.
https://api.tourmind.com/mcp/tocclaude mcp add --transport http tourmind-com-hotel-booking-ai https://api.tourmind.com/mcp/toc{"mcpServers":{"tourmind-com-hotel-booking-ai":{"type":"http","url":"https://api.tourmind.com/mcp/toc"}}}Open in Cursor{"servers":{"tourmind-com-hotel-booking-ai":{"type":"http","url":"https://api.tourmind.com/mcp/toc"}}}Open in VS Code{"mcpServers":{"tourmind-com-hotel-booking-ai":{"type":"http","url":"https://api.tourmind.com/mcp/toc"}}}{"mcpServers":{"tourmind-com-hotel-booking-ai":{"serverUrl":"https://api.tourmind.com/mcp/toc"}}}Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.
Tools
11| Parameter | Type | Required | Description |
|---|---|---|---|
| adults | integer | No | Number of adults in each room |
| children | integer | No | Number of children in each room |
| user_key | string | No | Optional TourMind user key for authenticated ToC access |
| hotel_ids | null|array | Yes | TourMind hotel identifiers to query |
| room_count | integer | No | Number of rooms using the same occupancy |
| check_in_date | string | No | Check-in date in YYYY-MM-DD format |
| children_ages | null|array | No | Ages of the children in each room, each from 0 to 17 |
| check_out_date | string | No | Check-out date in YYYY-MM-DD format |
Details
From the project READMEHotel Booking AI MCP Product Package
This repository contains the ToC product contract and the user-installable companion Skill for Hotel Booking AI MCP. It intentionally does not contain MCP server source code, deployment scripts, or runtime-specific implementation.
Package contents
Hotel-Booking-AI-MCP/
├── README.md
├── Hotel Booking AI MCP-FORMAT.md
└── skill/
└── hotel-booking-ai/
├── SKILL.md
└── references/
└── parameter_guide.md
skill/hotel-booking-ai/defines the Agent's hotel-search, live-rate, booking, payment, order-management, and update behavior.Hotel Booking AI MCP-FORMAT.mddefines the public MCP connection and capability contract.
MCP connection
Use the public production endpoint without connection headers:
{
"mcpServers": {
"tourmind": {
"url": "https://api.tourmind.com/mcp/toc",
"type": "streamableHttp"
}
}
}
The MCP connection itself is public. check_skill_update, hotel search, hotel detail, single-hotel rates, batch rates, and availability checks require no user_key. Booking, order lookup, cancellation, and payment tools receive user_key from the companion Skill only when those protected actions are requested.
Product model
The ToC MCP and ToB MCP use the same hotel discovery, Google POI resolution, candidate verification, live-rate ranking, image, cancellation, tax, payment, and error-handling prompts wherever their product behavior is identical.
The ToC variant differs only where the product differs:
| Concern | ToC behavior |
|---|---|
| MCP endpoint | https://api.tourmind.com/mcp/toc |
| Connection authentication | Public connection; no headers |
| Discovery and availability | Public; no user_key required |
| Booking and order operations | Require user_key from user_key.txt |
Read-only web_url | Optional anonymous result returned by hotel-search or room-rate tools; no sign-in required; the companion Skill localizes the result-page URL to the user's language |
The MCP service provides stable hotel tools, including batch_query_room_rates for partial-success live-rate retrieval across up to 20 hotels. Search, single/batch rate, availability, and booking tools support repeated per-room adult/child occupancy. The local Skill defines the current user workflow and can evolve independently as search, ranking, price verification, booking, and display strategies improve.
Skill version and updates
The installed SKILL.md frontmatter is the single source of truth for the Skill version:
metadata:
author: TourMind
version: "<current-version>"
The Agent sends metadata.version only to check_skill_update when an update check is due. Hotel and order tools never receive a Skill version, and users never maintain it in the MCP connection configuration.
Development handoff
The server implementation must satisfy Hotel Booking AI MCP-FORMAT.md, including all eleven tools referenced by the companion Skill. Transport framework, hosting, deployment, internal forwarding, and release infrastructure are outside this package.
TourMind hotel booking ecosystem
Choose the package that matches the audience and connection model:
| Audience | Integration | Authentication model | Repository |
|---|---|---|---|
| Consumer / ToC | Direct HTTP Skill | Public search and availability; user_key only for order operations | Hotel Booking AI |
| Business / ToB | Direct HTTP Skill | Skill Token required for every API call | TourMind Booking Skill |
| Consumer / ToC | MCP package + companion Skill | Public MCP connection; user_key only for order operations | Hotel Booking AI MCP |
| Business / ToB | MCP package + companion Skill | Bearer-authenticated MCP connection | TourMind Booking MCP |
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.