{"@context":"https://schema.org","@type":"NewsArticle","generatedAt":"2026-08-11T09:21:12.743Z","headline":"MCP 无状态更新：如何扩展 AI 智能体基础设施","description":"2026-07-28 发布的 Model Context Protocol （MCP） 规范以完全无状态核心取代传统有状态约束，支持云原生水平扩展、无服务器部署和标准轮询负载均衡。","url":"https://www.aioga.com/news/cmsgg4c4504o6ro5qnxomqial/","mainEntityOfPage":"https://www.aioga.com/news/cmsgg4c4504o6ro5qnxomqial/","datePublished":"2026-08-05T18:52:46.669Z","dateModified":"2026-08-05T18:52:46.669Z","inLanguage":"zh-CN","publisher":{"@type":"NewsMediaOrganization","name":"Aioga","url":"https://www.aioga.com"},"citation":["https://developers.googleblog.com/scaling-ai-agent-infrastructure-with-the-mcp-stateless-updates","https://aihot.virxact.com/items/cmsgg4c4504o6ro5qnxomqial"],"canonicalUrl":"https://www.aioga.com/news/cmsgg4c4504o6ro5qnxomqial/","directAnswer":{"@type":"Answer","text":"Google Developers Blog称，2026年7月28日的MCP规范候选版本移除了传输层会话管理，以完全无状态的协议核心支持普通HTTP负载均衡、云原生水平扩展和无服务器部署。","url":"https://www.aioga.com/news/cmsgg4c4504o6ro5qnxomqial/","dateCreated":"2026-08-05T18:52:46.669Z","author":{"@type":"Organization","@id":"https://www.aioga.com/authors/aioga-editorial/#editorial-team","name":"Aioga Editorial Team","url":"https://www.aioga.com/authors/aioga-editorial/"}},"evidence":[{"@type":"CreativeWork","name":"developers.googleblog.com source article","url":"https://developers.googleblog.com/scaling-ai-agent-infrastructure-with-the-mcp-stateless-updates","datePublished":"2026-08-05T18:52:46.669Z","provider":{"@type":"Organization","name":"developers.googleblog.com","url":"https://developers.googleblog.com/scaling-ai-agent-infrastructure-with-the-mcp-stateless-updates"}},{"@type":"CreativeWork","name":"AIHot archive record","url":"https://aihot.virxact.com/items/cmsgg4c4504o6ro5qnxomqial","datePublished":"2026-08-05T18:52:46.669Z","provider":{"@type":"Organization","name":"AIHot","url":"https://aihot.virxact.com/items/cmsgg4c4504o6ro5qnxomqial"}}],"aggregationSource":"Google Developers Blog（RSS）","originalPublisher":{"name":"developers.googleblog.com","url":"https://developers.googleblog.com/scaling-ai-agent-infrastructure-with-the-mcp-stateless-updates"},"geoDeepAnswer":null,"article":{"id":"cmsgg4c4504o6ro5qnxomqial","slug":"cmsgg4c4504o6ro5qnxomqial","url":"https://www.aioga.com/news/cmsgg4c4504o6ro5qnxomqial/","title":"MCP 无状态更新：如何扩展 AI 智能体基础设施","title_en":"Scaling AI Agent Infrastructure with the MCP Stateless updates","summary":"2026-07-28 发布的 Model Context Protocol （MCP） 规范以完全无状态核心取代传统有状态约束，支持云原生水平扩展、无服务器部署和标准轮询负载均衡。","source":"Google Developers Blog（RSS）","sourceUrl":"https://developers.googleblog.com/scaling-ai-agent-infrastructure-with-the-mcp-stateless-updates","aiHotUrl":"https://aihot.virxact.com/items/cmsgg4c4504o6ro5qnxomqial","publishedAt":"2026-08-05T18:52:46.669Z","category":"行业动态","score":59,"selected":false,"articleBody":["As you deploy agentic workflows and scale up your users, your bottlenecks change. When the Model Context Protocol (MCP) was first introduced in late 2024, it provided an elegant, session-oriented framework that allowed LLMs to negotiate capabilities, invoke external tools, and retrieve contextual resources. It was perfect for a single client talking to a single server on a local machine and optimized for stdio.","But when we at Google began deploying MCP servers across our cloud-native infrastructure, we hit a hard wall. The original protocol-level session model required persistent state, handshakes, and session pinning. In short, it was built on stateful transports that broke the core tenets of modern cloud-native scalability.","To solve this, Google led the charge：https://github.com/modelcontextprotocol/modelcontextprotocol/pull/2575 to decouple the protocol from stateful transport constraints. Our teams needed MCP to scale across millions of concurrent queries on Google Cloud, and we knew that you all needed MCP to be ready for real-world enterprise scale too. Working closely with Hugging Face and other industry partners, we co-founded the MCP Transports Working Group.","Today, we are thrilled to celebrate the culmination of that work: the 2026-07-28 Model Context Protocol specification release candidate , which is already being widely adopted：https://aaif.io/blog/the-ecosystem-responds-to-stateless-mcp. This landmark release removes transport-level session management entirely, giving you a stateless protocol core that scales on ordinary HTTP load-balanced infrastructure. It’s the biggest change to MCP spec since launch, and if you don’t read the rest of this article, rest assured, it’s a change for the better - more scale, more secure, just as easy .","In the original protocol model (specification version 2025-11-25) [392], connecting to an MCP server over HTTP required a stateful initialization process:","The server responded with an Mcp-Session-Id header. To make any subsequent tool call or resource query, the client had to include that unique session ID on every request, pinning the client to the specific container or pod that held its in-memory session state.","This stateful constraint breaks the horizontal scaling models that cloud-native engineers depend on:","The new 2026-07-28 specification solves this by making the protocol core completely stateless. The handshake is gone . The initialize / initialized handshake (SEP-2575) and the logical Mcp-Session-Id header (SEP-2567) have been removed entirely.","Instead, every request is now self-describing and independent . Protocol version, client info, and client capabilities that used to be exchanged once at connection setup now travel in a _meta field inline on every single request.","Here is how a stateless tool call looks under the new 2026-07-28 specification:","Without protocol sessions, we needed standard mechanisms to route and govern traffic efficiently. Working under the Transports Working Group, we helped design SEP-2243 (HTTP Standardization) [302, 542].","Streamable HTTP POST requests now carry specific HTTP headers:","These headers are mirrored to match the JSON-RPC body. If they disagree, the server rejects the request with a -32020 header mismatch code.","By promoting these values to standard HTTP headers, proxies, gateways, and load balancers can route, rate-limit, and audit traffic without inspecting the request body . For security and logging teams, this is a massive win that drastically lowers the latency and processing overhead at the gateway layer.","To eliminate the need for long-lived Server-Sent Events (SSE) connections just to monitor if a tool or prompt list changed, the spec introduces caching fields modeled after HTTP's Cache-Control. Tool and resource results can now return a ttlMs (Time-to-Live in milliseconds) and a cacheScope . Clients know exactly how long a tools/list response is fresh and whether it is safe to cache across multiple users.","One of the most complex challenges we faced in a stateless world was how to handle server-to-client requests. Under previous versions, if an MCP server needed user clarification (an \"elicitation prompt\") or a confirmation during a tool call, it had to keep an SSE connection open to push that request to the client.","Multi Round-Trip Requests (SEP-2322) solves this problem beautifully by restructuring the interaction lifecycle into self-contained steps:","Instead of blocking the thread or holding a connection open, the server immediately returns an InputRequiredResult with a requestState payload containing serialized context [398]:","The client prompts the user, gathers the boolean answer, and reissues the call with inputResponses and the echoed requestState . Because the requestState contains everything needed to resume the task, any server instance behind your load balancer can pick up the retry request !","Sometimes a tool call simply takes a long time to run. A database backup, a CRM sync, or a refund through a payment gateway can take anywhere from 10 to 60 seconds. Holding the client connection open blocks the customer conversation and creates massive connection queues.","The Tasks Extension graduates from an experimental feature to a robust, first-class protocol extension. Now, when a client calls a long-running tool, the server immediately returns a taskId and kicks off the execution in the background:","As the responsibility of managing state shifts from the transport layer to the application layer, security becomes paramount. The 2026-07-28 spec delivers several crucial security enhancements:","For the first time, MCP now has a formal deprecation policy. Features move through a structured Active -> Deprecated -> Removed lifecycle with a minimum 12-month transition window . Three features enter deprecation today:","All four Tier-1 SDKs (TypeScript, Python, Go, and C#) already have beta releases available supporting the 2026-07-28 specification. We highly encourage you to start testing these in your staging environments today.","In Python, the MCPServer decorator API is fully compatible [303]. You can install the beta directly with:","TypeScript v2 replaces the monolithic @modelcontextprotocol/sdk package with modular, focused libraries to keep your dependencies light. Install them with:","A convenient codemod is available to handle standard API renames (like renaming .tool() to registerTool ):","The 2026-07-28 specification marks a watershed moment for the Model Context Protocol, transitioning it from a promising local integration layer into the foundational, open infrastructure for enterprise AI applications.","Thank you to the huge effort from all of the MCP Transports Working Group：https://github.com/modelcontextprotocol/transports-wg/blob/main/GOVERNANCE.md#members and other teams who worked to make this happen, from across many companies. Thanks also to the Google team who maintain the Go MCP SDK：https://github.com/modelcontextprotocol/go-sdk and shipped v1.7.0：https://github.com/modelcontextprotocol/go-sdk/releases/tag/v1.7.0 on July 28th which was ready on launch day and powers major integrations like Github MCP Server：https://github.blog/changelog/2026-07-23-github-mcp-server-supports-the-next-mcp-specification/.","Google’s push for stateless transports was born out of necessity. We needed a protocol robust enough to handle the massive scale of our global developers, and we wanted to ensure that every developer, whether building on Google Cloud or anywhere else, had access to highly reliable, secure, and infinitely scalable agentic infrastructure.","By decoupling state from the transport layer, we have made load balancing boring, autoscaling seamless, and serverless deployment a reality. We can’t wait to see the incredibly scalable AI agents you build on top of this new stateless foundation!","Scaling real-time AI agents with session-aware load balancing","Agent and Model Evaluations in Gemini Enterprise Agent Platform are now GA"],"articleImages":[{"sourceUrl":"https://storage.googleapis.com/gweb-developer-goog-blog-cms-assets/site/20260731-000318/images/g-dev.svg","alt":"Google for Developers","afterParagraph":0,"url":"/media/articles/cmsgg4c4504o6ro5qnxomqial/0a905f88f3aab252.jpg"},{"sourceUrl":"https://storage.googleapis.com/gweb-developer-goog-blog-assets/images/Screenshot_2026-08-03_at_11.42.20AM.original.png","alt":"Screenshot 2026-08-03 at 11.42.20 AM","afterParagraph":6,"url":"/media/articles/cmsgg4c4504o6ro5qnxomqial/24ba57ba329f0b4b.png"},{"sourceUrl":"https://storage.googleapis.com/gweb-developer-goog-blog-assets/images/Screenshot_2026-08-03_at_11.51.27AM.original.png","alt":"Screenshot 2026-08-03 at 11.51.27 AM","afterParagraph":8,"url":"/media/articles/cmsgg4c4504o6ro5qnxomqial/7e0678337c5a2edf.png"},{"sourceUrl":"https://storage.googleapis.com/gweb-developer-goog-blog-assets/images/Screenshot_2026-08-03_at_11.44.18AM.original.png","alt":"Screenshot 2026-08-03 at 11.44.18 AM","afterParagraph":9,"url":"/media/articles/cmsgg4c4504o6ro5qnxomqial/342b65000f611b96.png"},{"sourceUrl":"https://storage.googleapis.com/gweb-developer-goog-blog-assets/images/Screenshot_2026-08-03_at_11.43.29AM_1.original.png","alt":"Screenshot 2026-08-03 at 11.43.29 AM (1)","afterParagraph":14,"url":"/media/articles/cmsgg4c4504o6ro5qnxomqial/f1f878e599324f39.png"},{"sourceUrl":"https://storage.googleapis.com/gweb-developer-goog-blog-assets/images/Screenshot_2026-08-03_at_11.44.30AM.original.png","alt":"Screenshot 2026-08-03 at 11.44.30 AM","afterParagraph":21,"url":"/media/articles/cmsgg4c4504o6ro5qnxomqial/eb5eaf0e9953aaf6.png"}],"mediaStatus":"ok","articleBodyZh":["随着您部署智能工作流并扩大用户规模，您的瓶颈也会发生变化。当模型上下文协议（MCP）在2024年底首次引入时，它提供了一个优雅的、面向会话的框架，使大型语言模型能够协商功能、调用外部工具并检索上下文资源。对于在本地机器上单个客户端与单个服务器的通信而言，它非常完美，并针对标准输入输出进行了优化。","但当我们在谷歌开始在云原生基础设施上部署MCP服务器时，我们遇到了一个严重的瓶颈。原始的协议级会话模型需要持久状态、握手以及会话固定。简而言之，它是建立在有状态传输之上的，这破坏了现代云原生可扩展性的核心原则。","为了解决这个问题，谷歌牵头推进：https://github.com/modelcontextprotocol/modelcontextprotocol/pull/2575，将协议与有状态传输约束解耦。我们的团队需要MCP能够在谷歌云上处理数百万并发查询，我们也知道你们都希望MCP能够为真实企业规模做好准备。与Hugging Face及其他行业合作伙伴密切合作，我们共同成立了MCP传输工作组。","今天，我们非常高兴地庆祝这项工作的成果：2026-07-28模型上下文协议规范发布候选版，该版本已被广泛采用：https://aaif.io/blog/the-ecosystem-responds-to-stateless-mcp。这一里程碑式的发布完全移除了传输级会话管理，为您提供一个无状态的协议核心，可在普通HTTP负载均衡基础设施上扩展。这是自MCP发布以来规范最大的变化，如果您不阅读本文其余部分，请放心，这是一项改进——更高的可扩展性、更安全、同样易用。","在原始协议模型（规范版本2025-11-25）[392]中，通过HTTP连接到MCP服务器需要一个有状态的初始化过程：","服务器会以Mcp-Session-Id头响应。要进行任何后续的工具调用或资源查询，客户端必须在每个请求中包含该唯一会话ID，将客户端固定到持有其内存会话状态的特定容器或Pod。","这一有状态限制破坏了云原生工程师依赖的水平扩展模型：","新的 2026-07-28 规范通过使协议核心完全无状态来解决这个问题。握手被取消。初始化/已初始化握手（SEP-2575）和逻辑 Mcp-Session-Id 头（SEP-2567）已被完全移除。","取而代之的是，现在每个请求都是自描述且独立的。协议版本、客户端信息以及以前在连接建立时只交换一次的客户端能力，现在都会作为 _meta 字段在每个请求中内联传输。","在新的 2026-07-28 规范下，无状态工具调用如下所示：","在没有协议会话的情况下，我们需要标准机制来高效路由和管理流量。在传输工作组（Transports Working Group）下工作时，我们帮助设计了 SEP-2243（HTTP 标准化）[302, 542]。","可流式 HTTP POST 请求现在携带特定的 HTTP 头：","这些头会与 JSON-RPC 主体进行镜像匹配。如果不一致，服务器会以 -32020 头不匹配代码拒绝请求。","通过将这些值提升为标准 HTTP 头，代理、网关和负载均衡器可以在不检查请求体的情况下进行路由、限流和审计流量。对于安全和日志团队来说，这是一个巨大的胜利，可以大幅降低网关层的延迟和处理开销。","为了消除仅为监控工具或提示列表变化而保持长连接的服务器发送事件（SSE）连接的需求，该规范引入了基于 HTTP Cache-Control 模型的缓存字段。工具和资源结果现在可以返回 ttlMs（以毫秒为单位的生存时间）和 cacheScope。客户端可以准确知道工具/列表响应的有效时间，以及是否可以跨多个用户安全缓存。","在无状态环境中，我们面临的最复杂挑战之一是如何处理服务器到客户端的请求。在以前的版本中，如果 MCP 服务器在工具调用期间需要用户澄清（\"引导提示\"）或确认，它必须保持 SSE 连接以将该请求推送给客户端。","多轮往返请求（SEP-2322）通过将交互生命周期重构为自包含步骤，完美地解决了这个问题：","服务器不会阻塞线程或保持连接开启，而是立即返回一个带有 requestState 负载、包含序列化上下文的 InputRequiredResult [398]：","客户端提示用户，收集布尔答案，并使用 inputResponses 以及回显的 requestState 重新发起调用。由于 requestState 包含恢复任务所需的所有信息，因此任何位于负载均衡器后的服务器实例都可以处理重试请求！","有时候，工具调用只是执行时间较长。数据库备份、CRM 同步或者通过支付网关进行退款可能需要 10 到 60 秒。保持客户端连接开启会阻塞客户对话，并造成大量连接排队。","Tasks 扩展从实验性功能升级为健全的一级协议扩展。现在，当客户端调用运行时间长的工具时，服务器会立即返回一个 taskId，并在后台启动执行：","随着管理状态的职责从传输层转移到应用层，安全性变得至关重要。2026-07-28 规范带来了几项关键的安全增强：","MCP 首次引入了正式的弃用政策。功能会经历结构化的 Active -> Deprecated -> Removed 生命周期，并有至少 12 个月的过渡期。今天有三个功能进入弃用阶段：","所有四个一级 SDK（TypeScript、Python、Go 和 C#）已经提供了支持 2026-07-28 规范的测试版。我们强烈建议您今天就在您的预发布环境中开始测试它们。","在 Python 中，MCPServer 装饰器 API 完全兼容 [303]。您可以直接使用以下命令安装测试版：","TypeScript v2 用模块化、针对性的小型库取代了单一的 @modelcontextprotocol/sdk 包，以保持您的依赖轻量化。安装方式如下：","提供了一个方便的 codemod 用于处理标准 API 重命名（例如将 .tool() 重命名为 registerTool）：","2026-07-28 规范标志着模型上下文协议的重要转折点，将其从一个有前景的本地集成层转变为企业 AI 应用的基础开放基础设施。","感谢所有 MCP 传输工作组的大力付出：https://github.com/modelcontextprotocol/transports-wg/blob/main/GOVERNANCE.md#members 以及来自多家公司的其他团队共同努力促成这一成果。同时感谢维护 Go MCP SDK 的 Google 团队：https://github.com/modelcontextprotocol/go-sdk，并于 7 月 28 日发布了 v1.7.0：https://github.com/modelcontextprotocol/go-sdk/releases/tag/v1.7.0，该版本在发布当天即可使用，并支持诸如 Github MCP Server 的主要集成：https://github.blog/changelog/2026-07-23-github-mcp-server-supports-the-next-mcp-specification/。","Google 推动无状态传输的初衷源于需求。我们需要一种能够应对全球开发者庞大规模的稳健协议，并希望确保每位开发者，无论是在 Google Cloud 上开发，还是在其他任何地方开发，都能使用高可靠性、安全且可无限扩展的智能基础设施。","通过将状态与传输层解耦，我们使负载均衡变得轻松，自动伸缩变得无缝，并使无服务器部署成为现实。我们迫不及待想看到大家在这一新的无状态基础上构建的极具可扩展性的 AI 代理！","通过会话感知负载均衡，扩展实时 AI 代理","Gemini 企业代理平台中的代理和模型评估现已 GA（全面可用）"],"translationStatus":"translated","bodyOrigin":"source-page","editorial":{"summary":"Google Developers Blog称，2026年7月28日的MCP规范候选版本移除了传输层会话管理，以完全无状态的协议核心支持普通HTTP负载均衡、云原生水平扩展和无服务器部署。","background":"MCP于2024年末推出，最初采用面向会话的模型。此前HTTP连接需要初始化握手并返回Mcp-Session-Id，后续请求须携带该标识，可能使客户端绑定保存会话状态的容器或Pod。","viewpoint":"Aioga判断，此次调整的核心价值是将协议能力与有状态传输约束解耦，而非增加新的智能体功能。它可能降低MCP服务接入现有云基础设施的复杂度，但材料未提供实际成本或性能数据。","implications":"握手和逻辑会话标识被移除后，MCP服务可能更容易使用标准轮询负载均衡并进行横向扩展。值得关注的是，应用层如仍需上下文或身份状态，仍可能需要自行设计相应机制。","nextStep":"建议开发者核对2026年7月28日候选规范中SEP-2575与SEP-2567的变更，并评估现有客户端、服务端及会话依赖。后续应关注正式规范状态、生态采用情况和迁移兼容说明。","evidenceRefs":["title","summary","articleBody","source"],"status":"published","aiGenerated":true,"autoApproved":true,"generatedBy":"aioga-editorial:gpt-5.6-sol","reviewedBy":"aioga-editorial-review:gpt-5.6-sol","generatedAt":"2026-08-05T19:43:44.623Z","sourceHash":"6349bc32408e5d5f","review":{"approved":true,"groundedness":95,"clarity":94,"duplicationRisk":18,"blockingIssues":[],"notes":["“应用层如仍需上下文或身份状态，仍可能需要自行设计相应机制”属于合理的技术推论，但来源材料未直接说明；当前已使用“如”“可能”等限定语，不构成阻断问题。","来源将该版本称为“release candidate”，候选内容已准确保留“规范候选版本”这一状态。"]},"validation":{"passed":true,"mode":"ai-auto","revisions":0,"checks":["schema","length","source-attribution","low-source-overlap","no-html","independent-ai-review"]}},"tags":["行业动态","Google Developers Blog（RSS）"],"translations":{"zh-CN":{"title":"MCP 无状态更新：如何扩展 AI 智能体基础设施","summary":"2026-07-28 发布的 Model Context Protocol （MCP） 规范以完全无状态核心取代传统有状态约束，支持云原生水平扩展、无服务器部署和标准轮询负载均衡。","category":"行业动态","source":"developers.googleblog.com","aggregationSource":"Google Developers Blog（RSS）","pageTitle":"MCP 无状态更新：如何扩展 AI 智能体基础设施 - Aioga AI资讯","description":"2026-07-28 发布的 Model Context Protocol （MCP） 规范以完全无状态核心取代传统有状态约束，支持云原生水平扩展、无服务器部署和标准轮询负载均衡。","url":"https://www.aioga.com/news/cmsgg4c4504o6ro5qnxomqial/","articleBody":["随着您部署智能工作流并扩大用户规模，您的瓶颈也会发生变化。当模型上下文协议（MCP）在2024年底首次引入时，它提供了一个优雅的、面向会话的框架，使大型语言模型能够协商功能、调用外部工具并检索上下文资源。对于在本地机器上单个客户端与单个服务器的通信而言，它非常完美，并针对标准输入输出进行了优化。","但当我们在谷歌开始在云原生基础设施上部署MCP服务器时，我们遇到了一个严重的瓶颈。原始的协议级会话模型需要持久状态、握手以及会话固定。简而言之，它是建立在有状态传输之上的，这破坏了现代云原生可扩展性的核心原则。","为了解决这个问题，谷歌牵头推进：https://github.com/modelcontextprotocol/modelcontextprotocol/pull/2575，将协议与有状态传输约束解耦。我们的团队需要MCP能够在谷歌云上处理数百万并发查询，我们也知道你们都希望MCP能够为真实企业规模做好准备。与Hugging Face及其他行业合作伙伴密切合作，我们共同成立了MCP传输工作组。","今天，我们非常高兴地庆祝这项工作的成果：2026-07-28模型上下文协议规范发布候选版，该版本已被广泛采用：https://aaif.io/blog/the-ecosystem-responds-to-stateless-mcp。这一里程碑式的发布完全移除了传输级会话管理，为您提供一个无状态的协议核心，可在普通HTTP负载均衡基础设施上扩展。这是自MCP发布以来规范最大的变化，如果您不阅读本文其余部分，请放心，这是一项改进——更高的可扩展性、更安全、同样易用。","在原始协议模型（规范版本2025-11-25）[392]中，通过HTTP连接到MCP服务器需要一个有状态的初始化过程：","服务器会以Mcp-Session-Id头响应。要进行任何后续的工具调用或资源查询，客户端必须在每个请求中包含该唯一会话ID，将客户端固定到持有其内存会话状态的特定容器或Pod。","这一有状态限制破坏了云原生工程师依赖的水平扩展模型：","新的 2026-07-28 规范通过使协议核心完全无状态来解决这个问题。握手被取消。初始化/已初始化握手（SEP-2575）和逻辑 Mcp-Session-Id 头（SEP-2567）已被完全移除。","取而代之的是，现在每个请求都是自描述且独立的。协议版本、客户端信息以及以前在连接建立时只交换一次的客户端能力，现在都会作为 _meta 字段在每个请求中内联传输。","在新的 2026-07-28 规范下，无状态工具调用如下所示：","在没有协议会话的情况下，我们需要标准机制来高效路由和管理流量。在传输工作组（Transports Working Group）下工作时，我们帮助设计了 SEP-2243（HTTP 标准化）[302, 542]。","可流式 HTTP POST 请求现在携带特定的 HTTP 头：","这些头会与 JSON-RPC 主体进行镜像匹配。如果不一致，服务器会以 -32020 头不匹配代码拒绝请求。","通过将这些值提升为标准 HTTP 头，代理、网关和负载均衡器可以在不检查请求体的情况下进行路由、限流和审计流量。对于安全和日志团队来说，这是一个巨大的胜利，可以大幅降低网关层的延迟和处理开销。","为了消除仅为监控工具或提示列表变化而保持长连接的服务器发送事件（SSE）连接的需求，该规范引入了基于 HTTP Cache-Control 模型的缓存字段。工具和资源结果现在可以返回 ttlMs（以毫秒为单位的生存时间）和 cacheScope。客户端可以准确知道工具/列表响应的有效时间，以及是否可以跨多个用户安全缓存。","在无状态环境中，我们面临的最复杂挑战之一是如何处理服务器到客户端的请求。在以前的版本中，如果 MCP 服务器在工具调用期间需要用户澄清（\"引导提示\"）或确认，它必须保持 SSE 连接以将该请求推送给客户端。","多轮往返请求（SEP-2322）通过将交互生命周期重构为自包含步骤，完美地解决了这个问题：","服务器不会阻塞线程或保持连接开启，而是立即返回一个带有 requestState 负载、包含序列化上下文的 InputRequiredResult [398]：","客户端提示用户，收集布尔答案，并使用 inputResponses 以及回显的 requestState 重新发起调用。由于 requestState 包含恢复任务所需的所有信息，因此任何位于负载均衡器后的服务器实例都可以处理重试请求！","有时候，工具调用只是执行时间较长。数据库备份、CRM 同步或者通过支付网关进行退款可能需要 10 到 60 秒。保持客户端连接开启会阻塞客户对话，并造成大量连接排队。","Tasks 扩展从实验性功能升级为健全的一级协议扩展。现在，当客户端调用运行时间长的工具时，服务器会立即返回一个 taskId，并在后台启动执行：","随着管理状态的职责从传输层转移到应用层，安全性变得至关重要。2026-07-28 规范带来了几项关键的安全增强：","MCP 首次引入了正式的弃用政策。功能会经历结构化的 Active -> Deprecated -> Removed 生命周期，并有至少 12 个月的过渡期。今天有三个功能进入弃用阶段：","所有四个一级 SDK（TypeScript、Python、Go 和 C#）已经提供了支持 2026-07-28 规范的测试版。我们强烈建议您今天就在您的预发布环境中开始测试它们。","在 Python 中，MCPServer 装饰器 API 完全兼容 [303]。您可以直接使用以下命令安装测试版：","TypeScript v2 用模块化、针对性的小型库取代了单一的 @modelcontextprotocol/sdk 包，以保持您的依赖轻量化。安装方式如下：","提供了一个方便的 codemod 用于处理标准 API 重命名（例如将 .tool() 重命名为 registerTool）：","2026-07-28 规范标志着模型上下文协议的重要转折点，将其从一个有前景的本地集成层转变为企业 AI 应用的基础开放基础设施。","感谢所有 MCP 传输工作组的大力付出：https://github.com/modelcontextprotocol/transports-wg/blob/main/GOVERNANCE.md#members 以及来自多家公司的其他团队共同努力促成这一成果。同时感谢维护 Go MCP SDK 的 Google 团队：https://github.com/modelcontextprotocol/go-sdk，并于 7 月 28 日发布了 v1.7.0：https://github.com/modelcontextprotocol/go-sdk/releases/tag/v1.7.0，该版本在发布当天即可使用，并支持诸如 Github MCP Server 的主要集成：https://github.blog/changelog/2026-07-23-github-mcp-server-supports-the-next-mcp-specification/。","Google 推动无状态传输的初衷源于需求。我们需要一种能够应对全球开发者庞大规模的稳健协议，并希望确保每位开发者，无论是在 Google Cloud 上开发，还是在其他任何地方开发，都能使用高可靠性、安全且可无限扩展的智能基础设施。","通过将状态与传输层解耦，我们使负载均衡变得轻松，自动伸缩变得无缝，并使无服务器部署成为现实。我们迫不及待想看到大家在这一新的无状态基础上构建的极具可扩展性的 AI 代理！","通过会话感知负载均衡，扩展实时 AI 代理","Gemini 企业代理平台中的代理和模型评估现已 GA（全面可用）"]},"en":{"title":"MCP Stateless Updates: How to Scale AI Agent Infrastructure","summary":"The Model Context Protocol (MCP) specification, released on 2026-07-28, replaces traditional stateful constraints with a fully stateless kernel, supporting cloud-native horizontal scaling, serverless deployment, and standard polling load balancing.","category":"Industry","source":"Google Developers Blog（RSS）","aggregationSource":"Google Developers Blog（RSS）","pageTitle":"MCP Stateless Updates: How to Scale AI Agent Infrastructure - Aioga AI News","description":"The Model Context Protocol (MCP) specification, released on 2026-07-28, replaces traditional stateful constraints with a fully stateless kernel, supporting cloud-native horizontal...","url":"https://www.aioga.com/en/news/cmsgg4c4504o6ro5qnxomqial/","contentTranslated":true,"sourceHash":"aa10aa557230d69c","translatedAt":"2026-08-05T19:24:06.456Z"},"ja":{"title":"MCPのステートレスアップデート:AIエージェントインフラの拡大方法","summary":"2026年7月28日に公開されたModel Context Protocol(MCP)仕様は、従来のステートフル制約を完全なステートレスカーネルに置き換え、クラウドネイティブの水平スケーリング、サーバーレス展開、標準ポーリングロードバランシングをサポートします。","category":"業界動向","source":"Google Developers Blog（RSS）","aggregationSource":"Google Developers Blog（RSS）","pageTitle":"MCPのステートレスアップデート:AIエージェントインフラの拡大方法 - Aioga AIニュース","description":"2026年7月28日に公開されたModel Context Protocol(MCP)仕様は、従来のステートフル制約を完全なステートレスカーネルに置き換え、クラウドネイティブの水平スケーリング、サーバーレス展開、標準ポーリングロードバランシングをサポートします。","url":"https://www.aioga.com/ja/news/cmsgg4c4504o6ro5qnxomqial/","contentTranslated":true,"sourceHash":"aa10aa557230d69c","translatedAt":"2026-08-05T19:24:06.172Z"},"ko":{"title":"MCP 상태 없는 업데이트: AI 에이전트 인프라 확장 방법","summary":"2026년 7월 28일에 발표된 모델 컨텍스트 프로토콜(MCP) 명세는 전통적인 상태 기반 제약 조건을 완전 상태 없는 커널로 대체하며, 클라우드 네이티브 수평 확장, 서버 없는 배포, 표준 폴링 부하 분산을 지원합니다.","category":"업계 동향","source":"Google Developers Blog（RSS）","aggregationSource":"Google Developers Blog（RSS）","pageTitle":"MCP 상태 없는 업데이트: AI 에이전트 인프라 확장 방법 - Aioga AI 뉴스","description":"2026년 7월 28일에 발표된 모델 컨텍스트 프로토콜(MCP) 명세는 전통적인 상태 기반 제약 조건을 완전 상태 없는 커널로 대체하며, 클라우드 네이티브 수평 확장, 서버 없는 배포, 표준 폴링 부하 분산을 지원합니다.","url":"https://www.aioga.com/ko/news/cmsgg4c4504o6ro5qnxomqial/","contentTranslated":true,"sourceHash":"aa10aa557230d69c","translatedAt":"2026-08-05T19:24:14.981Z"},"es":{"title":"Actualizaciones sin estado de MCP: Cómo escalar la infraestructura de agentes de IA","summary":"La especificación Model Context Protocol (MCP), publicada el 28-07-2026, reemplaza las restricciones con estado tradicionales por un núcleo totalmente sin estado, que soporta escalado horizontal nativo en la nube, despliegue serverless y balanceo estándar de carga de sondeo.","category":"Industria","source":"Google Developers Blog（RSS）","aggregationSource":"Google Developers Blog（RSS）","pageTitle":"Actualizaciones sin estado de MCP: Cómo escalar la infraestructura de agentes de IA - Aioga Noticias de IA","description":"La especificación Model Context Protocol (MCP), publicada el 28-07-2026, reemplaza las restricciones con estado tradicionales por un núcleo totalmente sin estado, que soporta escal...","url":"https://www.aioga.com/es/news/cmsgg4c4504o6ro5qnxomqial/","contentTranslated":true,"sourceHash":"aa10aa557230d69c","translatedAt":"2026-08-05T19:24:14.073Z"},"fr":{"title":"Mises à jour sans état MCP : comment évoluer l’infrastructure d’agents IA","summary":"La spécification du protocole de contexte modèle (MCP), publiée le 28-07-2026, remplace les contraintes traditionnelles avec état par un noyau entièrement sans état, prenant en charge l’évolutivité horizontale native cloud, le déploiement serverless et l’équilibrage standard de charge par sondage.","category":"Industrie","source":"Google Developers Blog（RSS）","aggregationSource":"Google Developers Blog（RSS）","pageTitle":"Mises à jour sans état MCP : comment évoluer l’infrastructure d’agents IA - Aioga Actualités IA","description":"La spécification du protocole de contexte modèle (MCP), publiée le 28-07-2026, remplace les contraintes traditionnelles avec état par un noyau entièrement sans état, prenant en cha...","url":"https://www.aioga.com/fr/news/cmsgg4c4504o6ro5qnxomqial/","contentTranslated":true,"sourceHash":"aa10aa557230d69c","translatedAt":"2026-08-05T19:24:23.548Z"},"de":{"title":"MCP Stateless Updates: Wie man die KI-Agenten-Infrastruktur skaliert","summary":"Die Model Context Protocol (MCP)-Spezifikation, veröffentlicht am 28.07.2026, ersetzt traditionelle zustandsbehaftete Einschränkungen durch einen vollständig zustandslosen Kernel, der cloud-native horizontale Skalierung, serverlose Bereitstellung und Standard-Polling-Lastverteilung unterstützt.","category":"行业动态","source":"Google Developers Blog（RSS）","aggregationSource":"Google Developers Blog（RSS）","pageTitle":"MCP Stateless Updates: Wie man die KI-Agenten-Infrastruktur skaliert - Aioga KI-News","description":"Die Model Context Protocol (MCP)-Spezifikation, veröffentlicht am 28.07.2026, ersetzt traditionelle zustandsbehaftete Einschränkungen durch einen vollständig zustandslosen Kernel,...","url":"https://www.aioga.com/de/news/cmsgg4c4504o6ro5qnxomqial/","contentTranslated":true,"sourceHash":"aa10aa557230d69c","translatedAt":"2026-08-05T19:24:22.196Z"},"pt-BR":{"title":"Atualizações Sem Estado do MCP: Como Escalar a Infraestrutura de Agentes de IA","summary":"A especificação Model Context Protocol (MCP), lançada em 28-07-2026, substitui as restrições tradicionais com estado por um kernel totalmente sem estado, suportando escalonamento horizontal nativo da nuvem, implantação serverless e balanceamento padrão de carga de polling.","category":"行业动态","source":"Google Developers Blog（RSS）","aggregationSource":"Google Developers Blog（RSS）","pageTitle":"Atualizações Sem Estado do MCP: Como Escalar a Infraestrutura de Agentes de IA - Aioga Notícias de IA","description":"A especificação Model Context Protocol (MCP), lançada em 28-07-2026, substitui as restrições tradicionais com estado por um kernel totalmente sem estado, suportando escalonamento h...","url":"https://www.aioga.com/pt-BR/news/cmsgg4c4504o6ro5qnxomqial/","contentTranslated":true,"sourceHash":"aa10aa557230d69c","translatedAt":"2026-08-05T19:24:30.569Z"},"ru":{"title":"Обновления MCP без статуса: как масштабировать инфраструктуру агентов ИИ","summary":"Спецификация Model Context Protocol (MCP), опубликованная 28.07.2026, заменяет традиционные ограничения состояния на полностью безсостоятельное ядро, поддерживая облачное горизонтальное масштабирование, безсерверное развертывание и стандартное балансирование нагрузки на опросы.","category":"行业动态","source":"Google Developers Blog（RSS）","aggregationSource":"Google Developers Blog（RSS）","pageTitle":"Обновления MCP без статуса: как масштабировать инфраструктуру агентов ИИ - Aioga Новости ИИ","description":"Спецификация Model Context Protocol (MCP), опубликованная 28.07.2026, заменяет традиционные ограничения состояния на полностью безсостоятельное ядро, поддерживая облачное горизонта...","url":"https://www.aioga.com/ru/news/cmsgg4c4504o6ro5qnxomqial/","contentTranslated":true,"sourceHash":"aa10aa557230d69c","translatedAt":"2026-08-05T19:24:31.747Z"},"ar":{"title":"تحديثات MCP بدون حالة: كيفية توسيع بنية وكلاء الذكاء الاصطناعي التحتية","summary":"تستبدل مواصفة بروتوكول السياق النموذجي (MCP)، التي صدرت في 2026-07-28، القيود التقليدية بالحالة بنواة بدون حالة بالكامل، تدعم التوسع الأفقي الأصلي للسحابة، والنشر بدون خادم، وموازنة حمل الاستطلاع القياسية.","category":"行业动态","source":"Google Developers Blog（RSS）","aggregationSource":"Google Developers Blog（RSS）","pageTitle":"تحديثات MCP بدون حالة: كيفية توسيع بنية وكلاء الذكاء الاصطناعي التحتية - Aioga أخبار الذكاء الاصطناعي","description":"تستبدل مواصفة بروتوكول السياق النموذجي (MCP)، التي صدرت في 2026-07-28، القيود التقليدية بالحالة بنواة بدون حالة بالكامل، تدعم التوسع الأفقي الأصلي للسحابة، والنشر بدون خادم، وموازن...","url":"https://www.aioga.com/ar/news/cmsgg4c4504o6ro5qnxomqial/","contentTranslated":true,"sourceHash":"aa10aa557230d69c","translatedAt":"2026-08-05T19:24:40.235Z"},"hi":{"title":"एमसीपी स्टेटलेस अपडेट: एआई एजेंट इंफ्रास्ट्रक्चर को कैसे स्केल करें","summary":"2026-07-28 को जारी मॉडल कॉन्टेक्स्ट प्रोटोकॉल (एमसीपी) विनिर्देश, पारंपरिक स्टेटफुल बाधाओं को पूरी तरह से स्टेटलेस कर्नेल से बदल देता है, जो क्लाउड-नेटिव हॉरिजॉन्टल स्केलिंग, सर्वर लेस डिप्लॉयमेंट और स्टैंडर्ड पोलिंग लोड बैलेंसिंग का समर्थन करता है।","category":"行业动态","source":"Google Developers Blog（RSS）","aggregationSource":"Google Developers Blog（RSS）","pageTitle":"एमसीपी स्टेटलेस अपडेट: एआई एजेंट इंफ्रास्ट्रक्चर को कैसे स्केल करें - Aioga AI समाचार","description":"2026-07-28 को जारी मॉडल कॉन्टेक्स्ट प्रोटोकॉल (एमसीपी) विनिर्देश, पारंपरिक स्टेटफुल बाधाओं को पूरी तरह से स्टेटलेस कर्नेल से बदल देता है, जो क्लाउड-नेटिव हॉरिजॉन्टल स्केलिंग, सर्वर...","url":"https://www.aioga.com/hi/news/cmsgg4c4504o6ro5qnxomqial/","contentTranslated":true,"sourceHash":"aa10aa557230d69c","translatedAt":"2026-08-05T19:24:38.793Z"},"it":{"title":"Aggiornamenti senza stato MCP: Come scalare l'infrastruttura degli agenti AI","summary":"La specifica Model Context Protocol (MCP), rilasciata il 28-07-2026, sostituisce i tradizionali vincoli con stato con un kernel completamente stateless, supportando la scala orizzontale cloud-nativa, il deployment serverless e il bilanciamento standard del carico tramite polling.","category":"行业动态","source":"Google Developers Blog（RSS）","aggregationSource":"Google Developers Blog（RSS）","pageTitle":"Aggiornamenti senza stato MCP: Come scalare l'infrastruttura degli agenti AI - Aioga Notizie IA","description":"La specifica Model Context Protocol (MCP), rilasciata il 28-07-2026, sostituisce i tradizionali vincoli con stato con un kernel completamente stateless, supportando la scala orizzo...","url":"https://www.aioga.com/it/news/cmsgg4c4504o6ro5qnxomqial/","contentTranslated":true,"sourceHash":"aa10aa557230d69c","translatedAt":"2026-08-05T19:24:48.752Z"},"nl":{"title":"MCP Stateless Updates: Hoe AI-agentinfrastructuur te schalen","summary":"De Model Context Protocol (MCP)-specificatie, uitgebracht op 28-07-2026, vervangt traditionele stateful constraints door een volledig stateless kernel, die cloud-native horizontale schaalbaarheid, serverless deployment en standaard polling-belastingbalansering ondersteunt.","category":"行业动态","source":"Google Developers Blog（RSS）","aggregationSource":"Google Developers Blog（RSS）","pageTitle":"MCP Stateless Updates: Hoe AI-agentinfrastructuur te schalen - Aioga AI-nieuws","description":"De Model Context Protocol (MCP)-specificatie, uitgebracht op 28-07-2026, vervangt traditionele stateful constraints door een volledig stateless kernel, die cloud-native horizontale...","url":"https://www.aioga.com/nl/news/cmsgg4c4504o6ro5qnxomqial/","contentTranslated":true,"sourceHash":"aa10aa557230d69c","translatedAt":"2026-08-05T19:24:48.717Z"},"tr":{"title":"MCP Durumsuz Güncellemeler: AI Ajanı Altyapısını Nasıl Ölçeklendirirler","summary":"2026-07-28 tarihinde yayımlanan Model Bağlam Protokolü (MCP) spesifikasyonu, geleneksel durum kısıtlamalarını tamamen durumsuz bir çekirdekle değiştirerek bulut tabanlı yatay ölçeklendirme, sunucusuz dağıtım ve standart anket yük dengelemesini destekler.","category":"行业动态","source":"Google Developers Blog（RSS）","aggregationSource":"Google Developers Blog（RSS）","pageTitle":"MCP Durumsuz Güncellemeler: AI Ajanı Altyapısını Nasıl Ölçeklendirirler - Aioga AI Haberleri","description":"2026-07-28 tarihinde yayımlanan Model Bağlam Protokolü (MCP) spesifikasyonu, geleneksel durum kısıtlamalarını tamamen durumsuz bir çekirdekle değiştirerek bulut tabanlı yatay ölçek...","url":"https://www.aioga.com/tr/news/cmsgg4c4504o6ro5qnxomqial/","contentTranslated":true,"sourceHash":"aa10aa557230d69c","translatedAt":"2026-08-05T19:24:56.120Z"},"vi":{"title":"Cập nhật MCP Stateless: Cách Mở rộng Hạ tầng Tác nhân AI","summary":"Đặc tả Giao thức Ngữ cảnh Mô hình (MCP), phát hành ngày 28-07-2026, thay thế các ràng buộc trạng thái truyền thống bằng kernel hoàn toàn không trạng thái, hỗ trợ mở rộng ngang trên nền tảng đám mây-gốc, triển khai không máy chủ và cân bằng tải kiểm tra tiêu chuẩn.","category":"行业动态","source":"Google Developers Blog（RSS）","aggregationSource":"Google Developers Blog（RSS）","pageTitle":"Cập nhật MCP Stateless: Cách Mở rộng Hạ tầng Tác nhân AI - Tin tức AI Aioga","description":"Đặc tả Giao thức Ngữ cảnh Mô hình (MCP), phát hành ngày 28-07-2026, thay thế các ràng buộc trạng thái truyền thống bằng kernel hoàn toàn không trạng thái, hỗ trợ mở rộng ngang trên...","url":"https://www.aioga.com/vi/news/cmsgg4c4504o6ro5qnxomqial/","contentTranslated":true,"sourceHash":"aa10aa557230d69c","translatedAt":"2026-08-05T19:24:56.482Z"},"id":{"title":"Pembaruan MCP Stateless: Cara Memperluas Infrastruktur Agen AI","summary":"Spesifikasi Model Context Protocol (MCP), yang dirilis pada 28-07-2026, menggantikan batasan stateful tradisional dengan kernel stateless penuh, mendukung skala horizontal cloud-native, deployment serverless, dan load balancing polling standar.","category":"行业动态","source":"Google Developers Blog（RSS）","aggregationSource":"Google Developers Blog（RSS）","pageTitle":"Pembaruan MCP Stateless: Cara Memperluas Infrastruktur Agen AI - Berita AI Aioga","description":"Spesifikasi Model Context Protocol (MCP), yang dirilis pada 28-07-2026, menggantikan batasan stateful tradisional dengan kernel stateless penuh, mendukung skala horizontal cloud-na...","url":"https://www.aioga.com/id/news/cmsgg4c4504o6ro5qnxomqial/","contentTranslated":true,"sourceHash":"aa10aa557230d69c","translatedAt":"2026-08-05T19:25:04.929Z"},"th":{"title":"อัปเดต MCP แบบไม่มีสถานะ: วิธีขยายโครงสร้างพื้นฐานเอเจนต์ AI","summary":"ข้อกําหนด Model Context Protocol (MCP) ซึ่งเผยแพร่เมื่อวันที่ 28 กรกฎาคม 2026 แทนที่ข้อจํากัดแบบ stateful แบบดั้งเดิมด้วยเคอร์เนลแบบ stateless เต็มรูปแบบ รองรับการขยายแนวนอนแบบคลาวด์เนทีฟ การติดตั้งแบบ serverless และการกระจายโหลดแบบ polling มาตรฐาน","category":"行业动态","source":"Google Developers Blog（RSS）","aggregationSource":"Google Developers Blog（RSS）","pageTitle":"อัปเดต MCP แบบไม่มีสถานะ: วิธีขยายโครงสร้างพื้นฐานเอเจนต์ AI - ข่าว AI Aioga","description":"ข้อกําหนด Model Context Protocol (MCP) ซึ่งเผยแพร่เมื่อวันที่ 28 กรกฎาคม 2026 แทนที่ข้อจํากัดแบบ stateful แบบดั้งเดิมด้วยเคอร์เนลแบบ stateless เต็มรูปแบบ รองรับการขยายแนวนอนแบบคลาว...","url":"https://www.aioga.com/th/news/cmsgg4c4504o6ro5qnxomqial/","contentTranslated":true,"sourceHash":"aa10aa557230d69c","translatedAt":"2026-08-05T19:25:05.062Z"},"pl":{"title":"Aktualizacje bezstanowe MCP: Jak skalować infrastrukturę agentów AI","summary":"Specyfikacja Model Context Protocol (MCP), wydana 28.07.2026, zastępuje tradycyjne ograniczenia stanowe w pełni bezstanowym jądrem, obsługując skalowanie poziome natywne w chmurze, wdrażanie serwerless oraz standardowe balansowanie obciążenia polling.","category":"行业动态","source":"Google Developers Blog（RSS）","aggregationSource":"Google Developers Blog（RSS）","pageTitle":"Aktualizacje bezstanowe MCP: Jak skalować infrastrukturę agentów AI - Aioga Wiadomości AI","description":"Specyfikacja Model Context Protocol (MCP), wydana 28.07.2026, zastępuje tradycyjne ograniczenia stanowe w pełni bezstanowym jądrem, obsługując skalowanie poziome natywne w chmurze,...","url":"https://www.aioga.com/pl/news/cmsgg4c4504o6ro5qnxomqial/","contentTranslated":true,"sourceHash":"aa10aa557230d69c","translatedAt":"2026-08-05T19:25:13.330Z"}}}}