发现优秀的 MCP 服务器

MCPM CLI

在 Claude App 中管理 MCP 服务器的命令行工具。

mcp-server-flyder

Flyder 的 MCP 服务器

MCP Continuity Server (Versão Simplificada)

MCP Continuity Server 兼容 SDK 1.7.0 的修正版本

Superargs

Please clarify what you mean by "AI MCP server args." I need more information to provide a helpful translation. Specifically, I need to understand: * **What is "AI MCP"?** Is it a specific software package, a general concept, or something else? Knowing what it refers to will help me understand the context of "server args." * **What kind of "server args" are you looking for?** Are you looking for command-line arguments, configuration file settings, or something else? * **What is the purpose of providing these arguments "during runtime"?** Are you trying to modify the server's behavior while it's running, or are you simply looking for the correct arguments to start the server with? Once I have a better understanding of what you're asking, I can provide a more accurate and relevant translation. For example, if "AI MCP" refers to a hypothetical "AI Model Control Platform" and you're looking for command-line arguments to specify the model to load at runtime, I might translate it as: **假设 "AI MCP" 指的是一个 "AI 模型控制平台"，并且您正在寻找命令行参数来指定在运行时加载的模型，那么我可以将其翻译为：** **在运行时提供 AI 模型控制平台 (AI MCP) 服务器参数。** (Zài yùnxíng shí tígōng AI móxíng kòngzhì píngtái (AI MCP) fúwùqì cānshù.) This translation assumes a specific meaning for "AI MCP" and "server args." Please provide more context so I can give you a more accurate translation.

Spring AI MCP Server Example

一个用于娱乐和实验的 Spring AI MCP 服务器示例项目。🚀 它使用带有虚拟 Person 数据的内存数据存储实现基本的 CRUD 操作。🤖

DeepMyst MCP Server

一个 MCP 服务器，允许您与 DeepMyst 进行对话，并使用我们的网关、路由器和 Token 优化在 LLM 之间进行协作。

MCP Server Requests

需要搭建一个 MCP 服务器吗？在这里提出请求。

Hedera Testnet Mirror Node MCP Server

镜子 (jìng zi)

Atlassian Data Center MCP

数据中心版本 Atlassian 产品 (Bitbucket, Confluence, JIRA) 的 MCP 服务器

SMS.ir MCP Server

短信宝 SMS.ir 消息服务的 MCP 服务器。

MCP Filesystem Server

镜子 (jìng zi)

mcp-server-exa-search

Zed 扩展，用于 Exa 的 MCP 服务器

hyper-mcp

一个快速、安全的 MCP 服务器，它通过 WebAssembly 插件扩展其功能。

mcp-server-weaviate

镜子 (jìng zi)

Azure Model Context Protocol (MCP) Hub

Okay, here's a breakdown of resources for building and integrating Model Context Protocol (MCP) servers on Azure using multiple languages. Since MCP is a relatively new and evolving area, direct, comprehensive "one-stop-shop" resources are still emerging. I'll provide the best available information, focusing on the core components and how to adapt them to different languages on Azure. **Understanding Model Context Protocol (MCP)** * **Core Concept:** MCP is designed to provide a standardized way for AI models to access contextual information (e.g., user data, environment data, session history) at runtime. This allows models to make more informed and personalized decisions. * **Key Components:** * **MCP Server:** The central component that manages and serves the contextual data. This is what you'll be building. * **MCP Client:** The code within your AI model or application that requests data from the MCP Server. * **Data Sources:** The systems that hold the contextual information (databases, APIs, caches, etc.). **General Approach for Building an MCP Server on Azure** 1. **Choose a Language/Framework:** Select a language and framework suitable for building a web API. Popular choices include: * **Python (with Flask or FastAPI):** Excellent for rapid development and has a large ecosystem of libraries. * **C# (.NET):** Strong performance, well-suited for enterprise applications, and integrates seamlessly with Azure services. * **Node.js (with Express):** Good for building scalable and real-time applications. * **Java (with Spring Boot):** Another robust option for enterprise-grade solutions. 2. **Define the MCP API:** Design the API endpoints that your MCP Server will expose. This will likely involve: * **Request Format:** How the client will request data (e.g., using a specific ID or set of parameters). JSON is a common choice. * **Response Format:** The structure of the data returned by the server (again, likely JSON). * **Authentication/Authorization:** How you'll secure the API to ensure only authorized clients can access the data. 3. **Implement the API Logic:** Write the code to: * Receive requests. * Fetch data from the appropriate data sources. * Transform the data into the required response format. * Handle errors gracefully. 4. **Deploy to Azure:** Choose an Azure service to host your MCP Server: * **Azure App Service:** A fully managed platform for hosting web applications. Good for most scenarios. * **Azure Functions:** Serverless compute, ideal for event-driven architectures or APIs with infrequent usage. * **Azure Kubernetes Service (AKS):** For more complex deployments requiring container orchestration. * **Azure Container Apps:** A serverless container service that simplifies deploying containerized applications. 5. **Secure the API:** Implement authentication and authorization. Options include: * **Azure Active Directory (Azure AD):** For enterprise identity management. * **API Keys:** A simpler approach for less sensitive data. * **Managed Identities:** Allow your Azure resources to authenticate to other Azure services without needing to manage credentials. 6. **Monitor and Log:** Use Azure Monitor to track the performance and health of your MCP Server. Implement logging to help diagnose issues. **Language-Specific Resources and Examples (Adaptable for MCP)** While direct MCP examples in multiple languages are scarce, you can adapt existing Azure API examples: * **Python (Flask/FastAPI):** * **Azure App Service with Python:** [https://learn.microsoft.com/en-us/azure/app-service/quickstart-python](https://learn.microsoft.com/en-us/azure/app-service/quickstart-python) * **Azure Functions with Python:** [https://learn.microsoft.com/en-us/azure/azure-functions/functions-create-first-function-python](https://learn.microsoft.com/en-us/azure/azure-functions/functions-create-first-function-python) * **Example (Conceptual):** You would adapt these examples to: * Define API endpoints for your MCP data requests (e.g., `/context/{user_id}`). * Fetch data from your data sources (e.g., Azure Cosmos DB, Azure SQL Database). * Return the data in a JSON format. * **C# (.NET):** * **Azure App Service with .NET:** [https://learn.microsoft.com/en-us/azure/app-service/quickstart-dotnetcore](https://learn.microsoft.com/en-us/azure/app-service/quickstart-dotnetcore) * **Azure Functions with C#:** [https://learn.microsoft.com/en-us/azure/azure-functions/functions-create-first-function-vs](https://learn.microsoft.com/en-us/azure/azure-functions/functions-create-first-function-vs) * **Example (Conceptual):** Similar to Python, you'd create API controllers to handle MCP requests, access data sources using Entity Framework or other data access libraries, and return JSON responses. * **Node.js (Express):** * **Azure App Service with Node.js:** [https://learn.microsoft.com/en-us/azure/app-service/quickstart-nodejs](https://learn.microsoft.com/en-us/azure/app-service/quickstart-nodejs) * **Azure Functions with Node.js:** [https://learn.microsoft.com/en-us/azure/azure-functions/functions-create-first-function-node](https://learn.microsoft.com/en-us/azure/azure-functions/functions-create-first-function-node) * **Example (Conceptual):** Use Express to define routes for your MCP API, connect to data sources using libraries like `pg` (for PostgreSQL) or `mongodb` (for MongoDB), and return JSON data. * **Java (Spring Boot):** * **Azure App Service with Java:** [https://learn.microsoft.com/en-us/azure/app-service/quickstart-java](https://learn.microsoft.com/en-us/azure/app-service/quickstart-java) * **Azure Functions with Java:** [https://learn.microsoft.com/en-us/azure/azure-functions/functions-create-first-java](https://learn.microsoft.com/en-us/azure/azure-functions/functions-create-first-java) * **Example (Conceptual):** Use Spring Boot's REST controller features to create API endpoints, use Spring Data JPA or JDBC to access databases, and return JSON responses. **Key Considerations for MCP Implementation** * **Data Consistency:** Ensure that the data served by your MCP Server is consistent and up-to-date. Consider using caching mechanisms (e.g., Azure Cache for Redis) to improve performance and reduce load on your data sources. * **Scalability:** Design your MCP Server to handle a large number of requests. Azure App Service and Azure Functions can scale automatically. For more demanding workloads, consider AKS. * **Latency:** Minimize the latency of data retrieval. Optimize your data queries and use caching effectively. Consider the geographic location of your MCP Server and your AI models. * **Security:** Protect the data served by your MCP Server. Use strong authentication and authorization mechanisms. Encrypt data in transit and at rest. * **Data Governance:** Implement policies to ensure that data is used responsibly and ethically. Comply with relevant data privacy regulations (e.g., GDPR, CCPA). **Example Scenario (Python with FastAPI on Azure App Service)** 1. **FastAPI App:** ```python from fastapi import FastAPI, HTTPException from azure.cosmos import CosmosClient, PartitionKey from typing import Optional import os app = FastAPI() # Azure Cosmos DB Configuration (replace with your actual values) COSMOS_ENDPOINT = os.environ["COSMOS_ENDPOINT"] COSMOS_KEY = os.environ["COSMOS_KEY"] DATABASE_NAME = "mcp_db" CONTAINER_NAME = "user_context" # Initialize Cosmos DB client cosmos_client = CosmosClient(COSMOS_ENDPOINT, COSMOS_KEY) database = cosmos_client.get_database_client(DATABASE_NAME) container = database.get_container_client(CONTAINER_NAME) @app.get("/context/{user_id}") async def get_user_context(user_id: str): """ Retrieves user context data from Cosmos DB. """ try: item = container.read_item(item=user_id, partition_key=user_id) return item except Exception as e: raise HTTPException(status_code=404, detail="User context not found") @app.get("/health") async def health_check(): return {"status": "ok"} ``` 2. **Deployment to Azure App Service:** * Create an Azure App Service instance. * Configure environment variables for `COSMOS_ENDPOINT` and `COSMOS_KEY`. * Deploy the Python code to the App Service. You'll likely need a `requirements.txt` file listing dependencies (e.g., `fastapi`, `azure-cosmos`). **Important Notes:** * **MCP is Evolving:** The Model Context Protocol is still under development. Expect changes and updates to the specifications and available tools. * **Customization is Key:** You'll need to tailor your MCP Server to the specific needs of your AI models and data sources. * **Security Best Practices:** Always prioritize security when building and deploying your MCP Server. I hope this comprehensive guide helps you get started with building and integrating MCP servers on Azure using multiple languages! Remember to adapt the examples and resources to your specific requirements.

MCP Game Helper

自定义模型上下文协议 (MCP) 服务器，提供人工智能驱动的工具，以协助游戏开发者完成与战斗平衡、技能分析、关卡节奏和模拟相关的任务。

Git MCP Server Knowledge Base

Git MCP 服务器的实现、配置和故障排除的综合知识库

Docker Server Manager Go MCP

This looks like a combination of terms related to Docker, server management, Go programming language, and potentially a Media Control Platform (MCP). Here are a few possible translations, depending on the intended meaning: **Option 1 (Most Literal, assuming it's a project name):** * **Docker 服务器管理器 Go MCP:** (Docker Fúwùqì Guǎnlǐqì Go MCP) - This is a direct translation, keeping "MCP" as is. It's suitable if "MCP" is a well-known acronym within a specific context. **Option 2 (More Descriptive, if "MCP" is a Media Control Platform):** * **使用 Go 语言开发的 Docker 服务器媒体控制平台管理器:** (Shǐyòng Go yǔyán kāifā de Docker fúwùqì méitǐ kòngzhì píngtái guǎnlǐqì) - This translates to "Docker server media control platform manager developed using the Go language." This is more verbose but clarifies the purpose. **Option 3 (Slightly Shorter, if "MCP" is a Media Control Platform):** * **Go 语言 Docker 服务器媒体控制管理器:** (Go yǔyán Docker fúwùqì méitǐ kòngzhì guǎnlǐqì) - This translates to "Go language Docker server media control manager." It's a bit more concise than Option 2. **Option 4 (If "MCP" is a specific product or technology):** * **Docker 服务器管理器 Go [MCP 的中文名称]:** (Docker Fúwùqì Guǎnlǐqì Go [MCP de zhōngwén míngchēng]) - This is a placeholder. You would replace "[MCP 的中文名称]" with the actual Chinese name of the MCP product or technology. For example, if "MCP" refers to "Media Control Platform X," and "Media Control Platform X" is known as "媒体控制平台X (Méitǐ Kòngzhì Píngtái X)" in Chinese, then the translation would be: "Docker 服务器管理器 Go 媒体控制平台X" **Which translation is best depends on the context. To give you the *best* translation, I need more information about what "dockerServerMangerGoMCP" refers to. Specifically:** * **What is "MCP"?** Is it a specific product, a general type of platform, or something else? * **What is the purpose of this "dockerServerMangerGoMCP"?** What does it do? Once you provide more context, I can give you a much more accurate and helpful translation.

Cline Mcp Server Test

这是一个用于测试 Cline 和 MCP 服务器连接的仓库。

ms-365-mcp-server

Microsoft 365 MCP 服务器