Contentful MCP Remote Server
Exposes Contentful MCP tools as a remote HTTP endpoint, allowing Claude Desktop to interact with Contentful content management functionality via an Express.js server instead of running tools locally.
README
Contentful MCP Remote Server
This repository hosts an Express.js server designed to expose the Contentful MCP tools as a remote endpoint. It allows the Contentful MCP Desktop application to connect to and utilize the mcp-tools functionality via an HTTP API, rather than running them locally.
🚀 Getting Started (Remote Server)
These instructions will get your remote MCP server up and running.
Prerequisites
-
Node.js (LTS version recommended)
-
npm or Yarn
Installation
-
Clone the repository:
git clone [your-repository-url] cd [your-repository-name] -
Install dependencies:
npm install # or yarn install
Running the Server
To start the Express.js server:
npm start
# or
yarn start
The server will typically run on http://localhost:3000 (or another port if configured) and expose the MCP endpoint at /mcp. Ensure this server is accessible from where you're running the Contentful MCP Desktop application (e.g., if deploying to a cloud service, make sure the URL is public).
🔌 Connecting to the Contentful MCP Desktop App
To connect your Contentful MCP Desktop application to this remote server, you'll need to configure it to use mcp-remote globally and then point it to your server's endpoint.
1. Install mcp-remote globally
First, ensure you have mcp-remote installed globally on your machine. This command-line tool acts as a bridge between the desktop app and your remote server.
npm install -g mcp-remote
# or
yarn global add mcp-remote
2. Configure Claude Desktop for Remote Server
Open your Claude Desktop application's configuration file. This file is usually located in your user's application data directory. Add the following JSON snippet to the mcpServers section:
{
"mcpServers": {
"contentful_mcp_server": {
"command": "npx",
"args": [
"mcp-remote",
"https://example.com/mcp"
]
}
}
}
Important: Replace https://example.com/mcp with the actual URL where your Express.js server's /mcp endpoint is accessible.
After saving the configuration, restart the Claude Desktop application. You should now see "contentful_mcp_server" available as an option for running MCP commands.
3. Verify MCP Connection (Optional)
You can verify that your MCP server is correctly exposed and accessible by using the @modelcontextprotocol/inspector tool.
npx @modelcontextprotocol/inspector
This command will attempt to connect to the configured MCP servers and report their status, helping you confirm that your remote or local server is properly recognized.
💻 Running MCP Tools Locally (Alternative Setup)
If you prefer to run the Contentful MCP tools directly on your local machine without the remote server, you can follow these instructions.
1. Clone and Run the Base Repository
-
Clone the original Contentful MCP Server repository:
git clone https://github.com/applydigital/contentful-express-mcp-server cd contentful-express-mcp-server -
Install dependencies:
npm install # or yarn install -
Start the local MCP server:
npm dev # or yarn devThis will typically start the MCP server on
http://localhost:3000.
2. Configure Claude Desktop for Local Server
Open your Claude Desktop application's configuration file. Add the following JSON snippet to the mcpServers section to connect to your locally running instance:
{
"mcpServers": {
"contentful_mcp_local": {
"command": "npx",
"args": [
"mcp-server",
"http://localhost:3000"
]
}
}
}
After saving the configuration, restart the Claude Desktop application. You should now see "contentful_mcp_local" available as an option for running MCP commands, which will execute directly against your local mcp-server instance.
💡 Considerations
-
Security: If deploying this server publicly, ensure you implement appropriate security measures (e.g., API keys, authentication, rate limiting) to protect your endpoint.
-
Scalability: For high-volume usage, consider deploying this server on a scalable platform (like Google Cloud Run, AWS Lambda, or similar serverless offerings) to handle concurrent requests efficiently.
-
Error Handling: Enhance the server's error handling to provide more informative responses in case of issues with
mcp-toolsexecution. -
Logging: Implement robust logging to monitor server activity and troubleshoot problems.
-
mcp-toolsDependency: This server assumes that the necessarymcp-toolsare available and correctly configured within its environment. Ensure that themcp-toolspackage is a dependency of this Express.js project. -
Network Latency: Running MCP tools remotely introduces network latency. For very frequent or performance-critical operations, a local setup might still be preferable.
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。