Hotel Booking MCP Server
Provides tools to check room availability, create bookings with automatic price calculation, and list bookings. Uses in-memory data with no authentication for demo purposes.
README
Hotel Booking MCP Server (No Auth)
A minimal, no-authentication MCP server for the hotel booking demo. Exposes three tools:
list_room_types— check availability, pricing, and amenities (optionally filtered by room type)create_booking— create a booking; automatically computes nights/total price and decrements availabilitylist_bookings— lists all bookings made so far (handy for demoing/debugging)
Data is in-memory and seeded with the same 4 room types from the build guide (Standard, Deluxe, Suite, Executive). It resets whenever the server restarts.
Deploy to Render (fastest free option)
- Create a free account at render.com if you don't have one.
- Push this folder to a new GitHub repo (or use Render's "Deploy from a public Git repo" — you can also zip-upload if you don't want to use Git; Render supports both).
- In Render, click New + → Web Service.
- Connect your repo (or paste the public repo URL).
- Settings:
- Runtime: Node
- Build Command:
npm install - Start Command:
node server.js - Instance Type: Free
- Click Create Web Service. Render will build and deploy — takes about 1–2 minutes.
- Once live, your MCP server URL will be:
That full URL (includinghttps://<your-service-name>.onrender.com/mcp/mcp) is what you'll register as the Server URL in Salesforce's Agentforce Registry.
Note: Render's free tier spins the service down after inactivity and takes ~30–60 seconds to wake up on the next request. For a live demo, hit the URL once a minute or two before you present so it's already warm.
Register in Salesforce
- Setup → Agentforce Registry → Registered MCP Servers → New.
- Server URL:
https://<your-service-name>.onrender.com/mcp - Authentication Method: No Authentication
- Create and Continue — no callback URL, no client ID/secret, no OAuth step at all.
- In Agentforce Builder, add the
list_room_typesandcreate_bookingtools to your agent's topics, same as previously planned.
Run locally (optional, to test before deploying)
npm install
node server.js
Server listens on port 3000 (or $PORT if set). Health check at GET /, MCP endpoint at POST /mcp.
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。