发现优秀的 MCP 服务器

GBox MCP Server

Gbox MCP 服务器

AverbePorto-MCP

AverbePorto MCP 服务器 (AverbePorto MCP fúwùqì)

Model-Context-Protocol Servers

我创建的所有 MCP 服务器

File Edit Check MCP Server

强制执行预读取检查和详细提交文档的 MCP 服务器

Mcp Server Amq

用于与 AWS AmazonMQ API 交互的 MCP 服务器

CHM to Markdown Converter

将 CHM 转换为 Markdown

Shell MCP Server

镜子 (jìng zi)

feishu-tools-mcp

MCP 服务器为 AI 编码代理提供飞书相关操作，例如 cursor 飞书 MCP 插件。

spring-mcp-server-sample

MCP服务器示例 (MCP fúwùqì shìlì)

Linear Remote MCP server

Linear 的远程模型上下文协议 (MCP) 服务器。 (Linear de yuǎnchéng móxíng shàngxiàwén xiéyì (MCP) fúwùqì.)

GenAIScript MCP Demo 🚀

GenAIScript 的 MCP 服务器功能演示

Bitwig MCP Server

Bitwig Studio 的 MCP 服务器

Nestjs Mcp

为你的 NestJS 应用集成 MCP 服务器

MCP Host Installation

一个安装其他 MCP 的 MCP。你将通过向你的 mcp.json 文件添加命令来手动安装最后一个 MCP。将此 MCP 添加到你最喜欢的主机，并让它安装你想要的任何服务器。

MCP_claude

这是为了演示如何为 Claude Desktop MCP 客户端构建一个 MCP 服务器。

cursor_agents

Okay, here are a few ways you could use an MCP (presumably referring to a Media Control Platform or similar system) server to add a team of experts into an agent flow, along with explanations and considerations: **Understanding the Goal** First, let's clarify what "adding a team of experts into the agent flow" means. This likely involves: * **Routing:** Directing specific types of customer interactions (calls, chats, emails) to the appropriate expert(s). * **Escalation:** Transferring an interaction from a general agent to an expert when the agent needs assistance. * **Collaboration:** Allowing agents to consult with experts in real-time (e.g., via chat, conference call) without transferring the customer. * **Knowledge Sharing:** Providing agents with access to expert knowledge bases or documentation. **Methods Using an MCP Server** Here are some common approaches, assuming your MCP server has capabilities like routing, presence management, and integration with other systems: **1. Skill-Based Routing (Most Common)** * **Concept:** Configure the MCP server to route interactions based on the skills required to handle them. The "experts" are defined as having specific skills. * **Implementation:** * **Skill Definition:** Define skills in the MCP server (e.g., "Product A Expert," "Technical Support - Level 2," "Spanish Language"). * **Expert Skill Assignment:** Assign the appropriate skills to each expert agent in the MCP system. * **Routing Rules:** Create routing rules that direct interactions with specific requirements (e.g., "Product A" inquiries) to agents with the "Product A Expert" skill. This often involves analyzing the customer's input (e.g., IVR selections, chat keywords, email subject) to determine the required skills. * **Queue Management:** The MCP server manages queues for each skill. If no experts are immediately available, the interaction is placed in the appropriate queue. * **Advantages:** Efficiently routes interactions to the right experts. Scalable as the team of experts grows. * **Considerations:** Requires accurate skill definition and assignment. Needs a mechanism to determine the required skills for each interaction (e.g., IVR, AI-powered intent analysis). **2. Presence-Based Routing** * **Concept:** Route interactions to experts based on their availability (presence status). * **Implementation:** * **Presence Integration:** The MCP server integrates with the expert's communication tools (e.g., softphone, chat client) to track their presence (available, busy, away, etc.). * **Routing Rules:** Create routing rules that only send interactions to experts who are currently available. * **Overflow Handling:** Define what happens if no experts are available (e.g., send to a general queue, offer a callback). * **Advantages:** Avoids sending interactions to unavailable experts. Improves customer experience. * **Considerations:** Requires reliable presence information. Needs a strategy for handling overflow situations. **3. Escalation/Transfer Functionality** * **Concept:** Allow general agents to transfer interactions to experts when they need assistance. * **Implementation:** * **Transfer Options:** Provide agents with a way to easily transfer interactions to specific experts or to a queue of experts. This might be a button in their agent desktop application. * **Warm Transfer vs. Cold Transfer:** Decide whether the agent should introduce the customer to the expert (warm transfer) or simply transfer the interaction without introduction (cold transfer). * **Context Transfer:** Ensure that relevant information about the interaction (e.g., customer history, previous interactions) is transferred along with the interaction. * **Advantages:** Allows agents to handle a wider range of issues. Provides access to expert knowledge when needed. * **Considerations:** Requires a user-friendly transfer interface. Needs a mechanism to ensure that context is transferred. **4. Collaboration Tools (Consultation)** * **Concept:** Enable agents to consult with experts in real-time without transferring the customer. * **Implementation:** * **Chat Integration:** Integrate a chat system into the agent desktop application that allows agents to communicate with experts. * **Conference Calling:** Allow agents to add experts to a conference call with the customer. * **Screen Sharing:** Enable agents to share their screen with experts for assistance. * **Advantages:** Allows agents to resolve complex issues quickly. Reduces the need for transfers. * **Considerations:** Requires a reliable communication platform. Needs a mechanism to manage expert availability. **5. Knowledge Base Integration** * **Concept:** Provide agents with access to a knowledge base that contains expert knowledge. * **Implementation:** * **Knowledge Base Platform:** Use a knowledge base platform to store and organize expert knowledge. * **Integration with Agent Desktop:** Integrate the knowledge base into the agent desktop application so that agents can easily search for information. * **AI-Powered Search:** Use AI to improve the accuracy and relevance of search results. * **Advantages:** Empowers agents to resolve issues independently. Reduces the need to consult with experts. * **Considerations:** Requires a well-maintained knowledge base. Needs a mechanism to ensure that the knowledge is accurate and up-to-date. **Example Scenario (Skill-Based Routing)** Let's say you have a team of experts who specialize in different products (Product A, Product B, Product C). 1. **Define Skills:** In your MCP server, define skills: "Product A Expert," "Product B Expert," "Product C Expert." 2. **Assign Skills:** Assign the appropriate skills to each expert agent. For example, Agent John might have the "Product A Expert" skill. 3. **Configure IVR:** In your IVR (Interactive Voice Response) system, ask the customer which product they need help with. 4. **Routing Rule:** Create a routing rule in the MCP server that says: "If the customer selects 'Product A' in the IVR, route the call to an agent with the 'Product A Expert' skill." 5. **Queue Management:** If no "Product A Expert" agents are available, the call is placed in a "Product A Expert" queue. **Chinese Translation of Key Terms** Here are some translations of key terms that might be helpful when discussing this with Chinese-speaking colleagues: * **MCP Server:** 媒体控制平台服务器 (Méitǐ Kòngzhì Píngtái Fúwùqì) * **Agent Flow:** 代理流程 (Dàilǐ Liúchéng) * **Team of Experts:** 专家团队 (Zhuānjiā Tuánduì) * **Skill-Based Routing:** 基于技能的路由 (Jīyú Jìnéng de Lùyóu) * **Presence-Based Routing:** 基于状态的路由 (Jīyú Zhuàngtài de Lùyóu) * **Escalation:** 升级 (Shēngjí) * **Transfer:** 转接 (Zhuǎnjiē) * **Collaboration:** 协作 (Xiézuò) * **Knowledge Base:** 知识库 (Zhīshì Kù) * **IVR (Interactive Voice Response):** 交互式语音应答 (Jiāohùshì Yǔyīn Yìngdá) * **Queue:** 队列 (Duìliè) * **Routing Rules:** 路由规则 (Lùyóu Guīzé) * **Agent Desktop:** 代理桌面 (Dàilǐ Zhuōmiàn) **Important Considerations** * **MCP Server Capabilities:** The specific features and capabilities of your MCP server will determine which methods are possible. Consult your MCP server documentation or vendor for details. * **Integration:** Integration with other systems (e.g., IVR, CRM, knowledge base) is crucial for many of these methods. * **Agent Training:** Ensure that agents are properly trained on how to use the new features and processes. * **Monitoring and Optimization:** Monitor the performance of the system and make adjustments as needed to optimize routing and efficiency. To give you more specific advice, please provide more details about your MCP server and the specific requirements of your agent flow. For example: * What is the name of your MCP server? * What features does it support (e.g., skill-based routing, presence management, API integration)? * What type of interactions are you handling (e.g., calls, chats, emails)? * What are the specific skills of your experts?

MCP Host Project

Okay, here's a translation of the English text "Showcases how to integrate Spring AI's support for MCP (Model Context Protocol) within Spring Boot applications, covering both server-side and client-side implementations." into Chinese: **Option 1 (More Literal):** 展示如何在 Spring Boot 应用中集成 Spring AI 对 MCP (模型上下文协议) 的支持，涵盖服务端和客户端的实现。 **Option 2 (Slightly More Natural):** 本示例展示如何在 Spring Boot 应用中整合 Spring AI 的 MCP (模型上下文协议) 支持，包括服务端和客户端的实现方式。 **Explanation of Choices:** * **展示 (zhǎnshì) / 本示例 (běn shìlì):** Both mean "showcase" or "demonstrate." "本示例" (this example) is slightly more common in technical documentation. * **集成 (jíchéng) / 整合 (zhěnghé):** Both mean "integrate." "整合" can sometimes imply a more thorough or seamless integration. * **Spring AI 对 MCP (模型上下文协议) 的支持 (Spring AI duì MCP (móxíng shàngxiàwén xiéyì) de zhīchí):** This is a direct translation of "Spring AI's support for MCP (Model Context Protocol)." The Chinese translation of "Model Context Protocol" is "模型上下文协议" (móxíng shàngxiàwén xiéyì). * **涵盖 (hāngài) / 包括 (bāokuò):** Both mean "covering" or "including." * **服务端 (fúwùduān) / 客户端 (kèhùduān):** These are standard translations for "server-side" and "client-side," respectively. * **实现 (shíxiàn) / 实现方式 (shíxiàn fāngshì):** Both mean "implementation." "实现方式" (implementation method/way) is slightly more descriptive. **Recommendation:** I would recommend **Option 2 (Slightly More Natural):** **本示例展示如何在 Spring Boot 应用中整合 Spring AI 的 MCP (模型上下文协议) 支持，包括服务端和客户端的实现方式。** This option sounds a bit more natural and is commonly used in Chinese technical documentation.

🤖 MCP Server — Agent X (Powered by Gemini Flash + Twitter API)

使用 MCP 服务器构建的 AI 代理

Bear MCP Server

镜子 (jìng zi)

Prometheus Alertmanager MCP Server

一个与 Prometheus Alertmanager 集成的模型上下文协议 (MCP) 服务器 (Yī gè yǔ Prometheus Alertmanager jíchéng de móxíng shàngxiàwén xiéyì (MCP) fúwùqì)

MCP Simple Server

一个简单的服务器，实现了用于文档搜索的模型上下文协议。

Modes MCP Server

镜子 (jìng zi)

Sample Model Context Protocol Demos

好的，这是关于如何将模型上下文协议与 AWS 结合使用的一些示例： **总览** 模型上下文协议 (Model Context Protocol, MCP) 是一种标准化方法，用于将上下文信息传递给机器学习模型。这使得模型能够根据手头的任务更好地理解和响应。在 AWS 环境中，MCP 可以与各种服务集成，以增强模型的性能和可解释性。 **使用场景和示例** 以下是一些使用 MCP 与 AWS 结合的常见场景和示例： * **个性化推荐:** * **场景:** 构建一个推荐系统，根据用户的历史行为、人口统计信息和当前上下文（例如，一天中的时间、位置）来推荐产品或内容。 * **如何使用 MCP:** * 将用户历史行为、人口统计信息和上下文信息编码为 MCP 格式。 * 将 MCP 数据传递给您的推荐模型（例如，在 Amazon SageMaker 上部署的模型）。 * 模型使用 MCP 数据来生成个性化推荐。 * **AWS 服务:** Amazon Personalize, Amazon SageMaker, Amazon DynamoDB (存储用户数据), Amazon Location Service (获取位置信息) * **示例代码 (伪代码):** ```python # 假设我们已经从 DynamoDB 获取了用户数据，并从 Location Service 获取了位置信息 user_id = "user123" user_data = get_user_data_from_dynamodb(user_id) location_data = get_location_data_from_location_service(user_id) # 构建 MCP 数据 mcp_data = { "user_id": user_id, "user_history": user_data["purchase_history"], "user_age": user_data["age"], "location": location_data["city"], "time_of_day": "evening" } # 将 MCP 数据传递给 SageMaker 模型 response = sagemaker_endpoint.invoke(mcp_data) # 从响应中获取推荐结果 recommendations = response["recommendations"] ``` * **欺诈检测:** * **场景:** 检测金融交易中的欺诈行为，考虑交易金额、交易地点、用户行为模式等上下文信息。 * **如何使用 MCP:** * 将交易金额、交易地点、用户行为模式等信息编码为 MCP 格式。 * 将 MCP 数据传递给您的欺诈检测模型。 * 模型使用 MCP 数据来评估交易的欺诈风险。 * **AWS 服务:** Amazon Fraud Detector, Amazon SageMaker, Amazon Kinesis (实时数据流), Amazon DynamoDB (存储用户行为数据) * **示例代码 (伪代码):** ```python # 假设我们已经从 Kinesis 获取了实时交易数据 transaction_data = get_transaction_data_from_kinesis() # 构建 MCP 数据 mcp_data = { "transaction_amount": transaction_data["amount"], "transaction_location": transaction_data["location"], "user_behavior": get_user_behavior_from_dynamodb(transaction_data["user_id"]) } # 将 MCP 数据传递给 Fraud Detector 或 SageMaker 模型 fraud_score = fraud_detector.detect_fraud(mcp_data) # 或 sagemaker_endpoint.invoke(mcp_data) # 根据欺诈评分采取行动 if fraud_score > threshold: flag_transaction_as_fraudulent() ``` * **自然语言处理 (NLP):** * **场景:** 提高文本分类、情感分析或机器翻译等 NLP 任务的准确性，通过提供文档的上下文信息（例如，作者、来源、主题）。 * **如何使用 MCP:** * 将文档内容、作者、来源、主题等信息编码为 MCP 格式。 * 将 MCP 数据传递给您的 NLP 模型。 * 模型使用 MCP 数据来更好地理解文本并提高性能。 * **AWS 服务:** Amazon Comprehend, Amazon SageMaker, Amazon S3 (存储文档), Amazon Kendra (企业搜索) * **示例代码 (伪代码):** ```python # 假设我们已经从 S3 获取了文档 document = get_document_from_s3("s3://my-bucket/my-document.txt") # 构建 MCP 数据 mcp_data = { "document_content": document, "document_author": "John Doe", "document_source": "News Article", "document_topic": "Politics" } # 将 MCP 数据传递给 Comprehend 或 SageMaker 模型 sentiment = comprehend.detect_sentiment(mcp_data) # 或 sagemaker_endpoint.invoke(mcp_data) # 分析情感 print(f"Sentiment: {sentiment}") ``` * **图像识别:** * **场景:** 提高图像识别的准确性，通过提供图像的上下文信息（例如，拍摄地点、时间、天气）。 * **如何使用 MCP:** * 将图像数据、拍摄地点、时间、天气等信息编码为 MCP 格式。 * 将 MCP 数据传递给您的图像识别模型。 * 模型使用 MCP 数据来更好地识别图像中的对象。 * **AWS 服务:** Amazon Rekognition, Amazon SageMaker, Amazon S3 (存储图像), Amazon Location Service (获取位置信息) * **示例代码 (伪代码):** ```python # 假设我们已经从 S3 获取了图像 image = get_image_from_s3("s3://my-bucket/my-image.jpg") # 获取图像的元数据 (例如，通过 EXIF 数据) image_metadata = get_image_metadata(image) # 构建 MCP 数据 mcp_data = { "image_data": image, "location": image_metadata["location"], "time": image_metadata["time"], "weather": get_weather_data_from_location_service(image_metadata["location"]) } # 将 MCP 数据传递给 Rekognition 或 SageMaker 模型 objects = rekognition.detect_objects(mcp_data) # 或 sagemaker_endpoint.invoke(mcp_data) # 识别图像中的对象 print(f"Objects detected: {objects}") ``` **关键考虑因素:** * **数据格式:** 定义清晰的 MCP 数据格式，以便模型能够正确解析和使用上下文信息。 可以使用 JSON 或其他结构化数据格式。 * **数据量:** 考虑 MCP 数据的大小和复杂性，以及它对模型性能的影响。 优化数据传输和处理流程。 * **安全性:** 确保 MCP 数据的安全性，特别是当包含敏感信息时。 使用 AWS Identity and Access Management (IAM) 控制访问权限。 * **模型训练:** 在训练模型时，使用包含上下文信息的 MCP 数据，以便模型能够学习如何利用这些信息。 * **模型部署:** 确保模型部署环境能够接收和处理 MCP 数据。 可以使用 Amazon SageMaker Endpoints 或其他部署选项。 * **监控:** 监控模型的性能，并根据需要调整 MCP 数据或模型参数。 使用 Amazon CloudWatch 监控指标。 **总结:** 模型上下文协议 (MCP) 是一种强大的工具，可以提高机器学习模型的性能和可解释性。 通过将 MCP 与 AWS 服务集成，您可以构建更智能、更个性化的应用程序。 以上示例展示了 MCP 在不同场景下的应用，并提供了使用 AWS 服务的指导。 请根据您的具体需求进行调整和扩展。 希望这些示例对您有所帮助！ 如果您有任何其他问题，请随时提出。