feat: align the MCP example and guides with the 2026-07-28 specification

- Rewrite the mcp-server example for the stateless 2026-07-28 spec: no
  initialize handshake, per-request _meta.protocolVersion validation, and
  the input_required round-trip via params.inputResponses (8 tests).
- Add bilingual migration guide guides/mcp/migrate-2026-07-28 covering the
  removals, deprecations, authorization hardening, and a migration checklist.
- Add the four official MCP/AAIF announcement sources to the reviewed catalog;
  regenerate catalog.json, README catalogs, and the content manifest (19 docs).

Author: xxg1413

README.md

@@ -68,14 +68,14 @@ which access or safety constraints deserve attention.
 
 <!-- catalog-index:start -->
 <!-- Generated by tools/render_readmes.py; edit catalog/ instead. -->
-> **29 reviewed resources** · Catalog reviewed 2026-09-02 · 21 intermediate or advanced · Primary sources first
+> **33 reviewed resources** · Catalog reviewed 2026-09-06 · 25 intermediate or advanced · Primary sources first
 
 ### Choose a learning path
 
 - [**Python foundations**](#path-foundations) — Learn the language, environments, dependencies, typing, and tests that reliable Python work depends on. (8 resources)
 - [**Web and APIs**](#path-web-apis) — Build typed services and applications that connect Python logic to users and other systems. (7 resources)
 - [**Automation**](#path-automation) — Turn repeatable work into maintainable scripts, browser workflows, and data pipelines. (7 resources)
-- [**AI agents**](#path-ai-agents) — Learn tools, structured output, state, evaluation, and the safety boundaries of agent systems. (7 resources)
+- [**AI agents**](#path-ai-agents) — Learn tools, structured output, state, evaluation, and the safety boundaries of agent systems. (11 resources)
 
 <a id="path-foundations"></a>
 ### Python foundations
@@ -137,6 +137,10 @@ Learn tools, structured output, state, evaluation, and the safety boundaries of
 | [Google Agent Development Kit](https://adk.dev/)<br><sub>Official docs</sub> | Google's official documentation for developing, evaluating, and deploying modular agent systems. | Advanced<br>English | API key typically required<br>Review permissions and side effects | 2026-08-31 |
 | [Instructor documentation](https://python.useinstructor.com/)<br><sub>Official project</sub> | Structured outputs and validation for LLMs powered by Pydantic, enabling predictable tool calling and schema enforcement. | Intermediate<br>English | API key typically required<br>Review permissions and side effects | 2026-09-02 |
 | [Model Context Protocol Python SDK](https://github.com/modelcontextprotocol/python-sdk)<br><sub>Official project</sub> | Official Python implementation of the Model Context Protocol for building MCP servers and clients. | Intermediate<br>English | No API key<br>Review permissions and side effects | 2026-09-02 |
+| [MCP 2026-07-28 specification release](https://blog.modelcontextprotocol.io/posts/2026-07-28/)<br><sub>Official docs</sub> | The official release announcement for the 2026-07-28 Model Context Protocol specification, covering the stateless redesign and breaking changes. | Intermediate<br>English | No API key<br>Low risk | 2026-09-06 |
+| [MCP 2026-07-28 release candidate overview](https://blog.modelcontextprotocol.io/posts/2026-07-28-release-candidate/)<br><sub>Official docs</sub> | The official release-candidate post that enumerates the 2026-07-28 breaking changes and the migration window for SDK maintainers. | Intermediate<br>English | No API key<br>Low risk | 2026-09-06 |
+| [Anthropic donating MCP to the Agentic AI Foundation](https://www.anthropic.com/news/donating-the-model-context-protocol-and-establishing-of-the-agentic-ai-foundation)<br><sub>Official docs</sub> | Anthropic's announcement establishing neutral Linux Foundation governance for MCP under the Agentic AI Foundation. | Intermediate<br>English | No API key<br>Low risk | 2026-09-06 |
+| [MCP under the Agentic AI Foundation](https://aaif.io/projects/model-context-protocol)<br><sub>Official project</sub> | The Agentic AI Foundation project page for MCP, the neutral governance home for specification development and roadmaps. | Intermediate<br>English | No API key<br>Low risk | 2026-09-06 |
 
 Missing an important official source? [Propose a resource or report a correction](https://github.com/flypythoncom/python/issues/new/choose).
 <!-- catalog-index:end -->

README_cn.md

@@ -62,14 +62,14 @@ python examples/product-slug/verify.py solution
 
 <!-- catalog-index:start -->
 <!-- Generated by tools/render_readmes.py; edit catalog/ instead. -->
-> **29 条已审核资源** · 最近整体审核：2026-09-02 · 21 条进阶或高级资源 · 一手来源优先
+> **33 条已审核资源** · 最近整体审核：2026-09-06 · 25 条进阶或高级资源 · 一手来源优先
 
 ### 选择学习路径
 
 - [**Python 基础**](#path-foundations) — 掌握语言基础、环境、依赖、类型与测试，为可靠的 Python 项目打好根基。 (8 条资源)
 - [**Web 与 API**](#path-web-apis) — 构建有类型约束的服务和应用，把 Python 逻辑连接到用户与其他系统。 (7 条资源)
 - [**自动化**](#path-automation) — 把重复工作变成可维护的脚本、浏览器工作流和数据管道。 (7 条资源)
-- [**AI Agent**](#path-ai-agents) — 理解工具调用、结构化输出、状态、评测以及 Agent 系统的安全边界。 (7 条资源)
+- [**AI Agent**](#path-ai-agents) — 理解工具调用、结构化输出、状态、评测以及 Agent 系统的安全边界。 (11 条资源)
 
 <a id="path-foundations"></a>
 ### Python 基础
@@ -131,6 +131,10 @@ python examples/product-slug/verify.py solution
 | [Google Agent Development Kit](https://adk.dev/)<br><sub>官方文档</sub> | Google 官方模块化 Agent 系统开发、评测与部署文档。 | 高级<br>英语 | 通常需要 API Key<br>检查权限与副作用 | 2026-08-31 |
 | [Instructor documentation](https://python.useinstructor.com/)<br><sub>官方项目</sub> | 基于 Pydantic 的大模型结构化输出与严格验证框架，确保工具调用与数据提取的确定性。 | 进阶<br>英语 | 通常需要 API Key<br>检查权限与副作用 | 2026-09-02 |
 | [Model Context Protocol Python SDK](https://github.com/modelcontextprotocol/python-sdk)<br><sub>官方项目</sub> | Model Context Protocol 官方 Python SDK，用于快速构建和集成标准的 MCP 工具与服务。 | 进阶<br>英语 | 无需 API Key<br>检查权限与副作用 | 2026-09-02 |
+| [MCP 2026-07-28 specification release](https://blog.modelcontextprotocol.io/posts/2026-07-28/)<br><sub>官方文档</sub> | Model Context Protocol 2026-07-28 规范的官方发布公告，说明无状态重构与破坏性变更。 | 进阶<br>英语 | 无需 API Key<br>低风险 | 2026-09-06 |
+| [MCP 2026-07-28 release candidate overview](https://blog.modelcontextprotocol.io/posts/2026-07-28-release-candidate/)<br><sub>官方文档</sub> | 官方 RC 发布说明，逐项列出 2026-07-28 的破坏性变更以及留给 SDK 维护者的迁移窗口。 | 进阶<br>英语 | 无需 API Key<br>低风险 | 2026-09-06 |
+| [Anthropic donating MCP to the Agentic AI Foundation](https://www.anthropic.com/news/donating-the-model-context-protocol-and-establishing-of-the-agentic-ai-foundation)<br><sub>官方文档</sub> | Anthropic 官方公告，宣布将 MCP 交由 Linux 基金会旗下的 Agentic AI Foundation 中立治理。 | 进阶<br>英语 | 无需 API Key<br>低风险 | 2026-09-06 |
+| [MCP under the Agentic AI Foundation](https://aaif.io/projects/model-context-protocol)<br><sub>官方项目</sub> | Agentic AI Foundation 的 MCP 项目主页，是规范开发与路线图的中立治理入口。 | 进阶<br>英语 | 无需 API Key<br>低风险 | 2026-09-06 |
 
 没有找到合适的官方资料？可以[建议资源或报告错误](https://github.com/flypythoncom/python/issues/new/choose)。
 <!-- catalog-index:end -->

catalog.json

@@ -2,7 +2,7 @@
   "$schema": "./schema/catalog-v1.schema.json",
   "schema_version": 1,
   "catalog": {
-    "reviewed_on": "2026-09-02",
+    "reviewed_on": "2026-09-06",
     "status": "active"
   },
   "paths": [
@@ -532,6 +532,74 @@
       "risk": "medium",
       "featured": false,
       "order": 7
+    },
+    {
+      "id": "mcp-2026-07-28-spec-release",
+      "path": "ai-agents",
+      "title": "MCP 2026-07-28 specification release",
+      "url": "https://blog.modelcontextprotocol.io/posts/2026-07-28/",
+      "source_type": "official-docs",
+      "level": "intermediate",
+      "language": "en",
+      "why_en": "The official release announcement for the 2026-07-28 Model Context Protocol specification, covering the stateless redesign and breaking changes.",
+      "why_zh": "Model Context Protocol 2026-07-28 规范的官方发布公告，说明无状态重构与破坏性变更。",
+      "reviewed_on": "2026-09-06",
+      "status": "active",
+      "requires_key": false,
+      "risk": "low",
+      "featured": false,
+      "order": 8
+    },
+    {
+      "id": "mcp-2026-07-28-release-candidate",
+      "path": "ai-agents",
+      "title": "MCP 2026-07-28 release candidate overview",
+      "url": "https://blog.modelcontextprotocol.io/posts/2026-07-28-release-candidate/",
+      "source_type": "official-docs",
+      "level": "intermediate",
+      "language": "en",
+      "why_en": "The official release-candidate post that enumerates the 2026-07-28 breaking changes and the migration window for SDK maintainers.",
+      "why_zh": "官方 RC 发布说明，逐项列出 2026-07-28 的破坏性变更以及留给 SDK 维护者的迁移窗口。",
+      "reviewed_on": "2026-09-06",
+      "status": "active",
+      "requires_key": false,
+      "risk": "low",
+      "featured": false,
+      "order": 9
+    },
+    {
+      "id": "anthropic-mcp-aaif-donation",
+      "path": "ai-agents",
+      "title": "Anthropic donating MCP to the Agentic AI Foundation",
+      "url": "https://www.anthropic.com/news/donating-the-model-context-protocol-and-establishing-of-the-agentic-ai-foundation",
+      "source_type": "official-docs",
+      "level": "intermediate",
+      "language": "en",
+      "why_en": "Anthropic's announcement establishing neutral Linux Foundation governance for MCP under the Agentic AI Foundation.",
+      "why_zh": "Anthropic 官方公告，宣布将 MCP 交由 Linux 基金会旗下的 Agentic AI Foundation 中立治理。",
+      "reviewed_on": "2026-09-06",
+      "status": "active",
+      "requires_key": false,
+      "risk": "low",
+      "featured": false,
+      "order": 10
+    },
+    {
+      "id": "aaif-model-context-protocol",
+      "path": "ai-agents",
+      "title": "MCP under the Agentic AI Foundation",
+      "url": "https://aaif.io/projects/model-context-protocol",
+      "source_type": "official-project",
+      "level": "intermediate",
+      "language": "en",
+      "why_en": "The Agentic AI Foundation project page for MCP, the neutral governance home for specification development and roadmaps.",
+      "why_zh": "Agentic AI Foundation 的 MCP 项目主页，是规范开发与路线图的中立治理入口。",
+      "reviewed_on": "2026-09-06",
+      "status": "active",
+      "requires_key": false,
+      "risk": "low",
+      "featured": false,
+      "order": 11
     }
   ]
 }

catalog/catalog.yml

@@ -1,2 +1,2 @@
-reviewed_on: '2026-09-02'
+reviewed_on: '2026-09-06'
 status: active

catalog/resources/aaif-model-context-protocol.yml

@@ -0,0 +1,15 @@
+id: aaif-model-context-protocol
+path: ai-agents
+title: MCP under the Agentic AI Foundation
+url: https://aaif.io/projects/model-context-protocol
+source_type: official-project
+level: intermediate
+language: en
+why_en: The Agentic AI Foundation project page for MCP, the neutral governance home for specification development and roadmaps.
+why_zh: Agentic AI Foundation 的 MCP 项目主页，是规范开发与路线图的中立治理入口。
+reviewed_on: '2026-09-06'
+status: active
+requires_key: false
+risk: low
+featured: false
+order: 11

catalog/resources/anthropic-mcp-aaif-donation.yml

@@ -0,0 +1,15 @@
+id: anthropic-mcp-aaif-donation
+path: ai-agents
+title: Anthropic donating MCP to the Agentic AI Foundation
+url: https://www.anthropic.com/news/donating-the-model-context-protocol-and-establishing-of-the-agentic-ai-foundation
+source_type: official-docs
+level: intermediate
+language: en
+why_en: Anthropic's announcement establishing neutral Linux Foundation governance for MCP under the Agentic AI Foundation.
+why_zh: Anthropic 官方公告，宣布将 MCP 交由 Linux 基金会旗下的 Agentic AI Foundation 中立治理。
+reviewed_on: '2026-09-06'
+status: active
+requires_key: false
+risk: low
+featured: false
+order: 10

catalog/resources/mcp-2026-07-28-release-candidate.yml

@@ -0,0 +1,15 @@
+id: mcp-2026-07-28-release-candidate
+path: ai-agents
+title: MCP 2026-07-28 release candidate overview
+url: https://blog.modelcontextprotocol.io/posts/2026-07-28-release-candidate/
+source_type: official-docs
+level: intermediate
+language: en
+why_en: The official release-candidate post that enumerates the 2026-07-28 breaking changes and the migration window for SDK maintainers.
+why_zh: 官方 RC 发布说明，逐项列出 2026-07-28 的破坏性变更以及留给 SDK 维护者的迁移窗口。
+reviewed_on: '2026-09-06'
+status: active
+requires_key: false
+risk: low
+featured: false
+order: 9

catalog/resources/mcp-2026-07-28-spec-release.yml

@@ -0,0 +1,15 @@
+id: mcp-2026-07-28-spec-release
+path: ai-agents
+title: MCP 2026-07-28 specification release
+url: https://blog.modelcontextprotocol.io/posts/2026-07-28/
+source_type: official-docs
+level: intermediate
+language: en
+why_en: The official release announcement for the 2026-07-28 Model Context Protocol specification, covering the stateless redesign and breaking changes.
+why_zh: Model Context Protocol 2026-07-28 规范的官方发布公告，说明无状态重构与破坏性变更。
+reviewed_on: '2026-09-06'
+status: active
+requires_key: false
+risk: low
+featured: false
+order: 8

content-manifest.json

@@ -73,23 +73,23 @@
     {
       "id": "example-mcp-server",
       "type": "example",
-      "content_version": 1,
+      "content_version": 2,
       "status": "reviewed",
-      "reviewed_on": "2026-09-02",
+      "reviewed_on": "2026-09-06",
       "locales": [
         {
           "lang": "en-US",
           "path": "examples/mcp-server/README.md",
           "title": "Model Context Protocol (MCP) Tool Server",
-          "summary": "Build a safe, compliant JSON-RPC 2.0 MCP tool server with parameter validation and structured error boundaries.",
-          "sha256": "12cbb64484ad5ccccb05d50566ef7c3c92eb9043641bf299b94b5247adaf12b2"
+          "summary": "Build a stateless 2026-07-28 MCP tool server with JSON-RPC 2.0 dispatch, no initialize handshake, schema validation, error isolation, and the input_required round-trip.",
+          "sha256": "989d76f36d383a73fde46c5e9a88df6efc6466b44f1b7a867c0745d1bdb1a0d6"
         },
         {
           "lang": "zh-CN",
           "path": "examples/mcp-server/README_cn.md",
           "title": "Model Context Protocol (MCP) 工具服务",
-          "summary": "构建规范安全的 JSON-RPC 2.0 MCP 工具服务，实现参数校验与结构化错误边界。",
-          "sha256": "ec1a70a7cb3304df5a1dd69412bf892f81dace748a7a3f56a811d3146c45b19f"
+          "summary": "构建符合 2026-07-28 无状态规范的 MCP 工具服务，实现无需 initialize 握手的 JSON-RPC 2.0 分发、参数校验、错误隔离与 input_required 多轮交互。",
+          "sha256": "e29cb70049b5bc5220226ab411b75c7a898242a061490a88fa1d7ef8ca7e9fa1"
         }
       ]
     },
@@ -185,6 +185,29 @@
         }
       ]
     },
+    {
+      "id": "mcp-2026-07-28-migration",
+      "type": "guide",
+      "content_version": 1,
+      "status": "reviewed",
+      "reviewed_on": "2026-09-06",
+      "locales": [
+        {
+          "lang": "en-US",
+          "path": "guides/mcp/migrate-2026-07-28.md",
+          "title": "Migrate a Python MCP server to the 2026-07-28 specification",
+          "summary": "Update a Python MCP server for the stateless 2026-07-28 specification by removing the initialize handshake, adopting the input_required round-trip, dropping deprecated features, and hardening authorization.",
+          "sha256": "47a96d64e7f2a86c39fcc716b1e65527961256ae44cb79e2817a13c3eb7feaa2"
+        },
+        {
+          "lang": "zh-CN",
+          "path": "guides/mcp/migrate-2026-07-28_cn.md",
+          "title": "将 Python MCP 服务端迁移到 2026-07-28 规范",
+          "summary": "面向 2026-07-28 无状态规范更新 Python MCP 服务端，移除 initialize 握手、采用 input_required 多轮交互、放弃废弃特性并强化授权。",
+          "sha256": "bb9eb48393dccc469b6ffcdd52248e82d71daf85dd9b891d00b3adf9b8027425"
+        }
+      ]
+    },
     {
       "id": "product-slug-example",
       "type": "example",

examples/README.md

@@ -6,7 +6,7 @@ loop without trusting generated code on sight.
 
 - [Product slug regression](product-slug/README.md): reproduce a text-boundary bug, ask an agent for the smallest fix, and verify the solution. Standard library only; about three minutes.
 - [Untrusted payload validation](pydantic-validation/README.md): handle messy camelCase and snake_case inputs, dirty amount strings, and structured error responses.
-- [MCP tool server](mcp-server/README.md): build a standard JSON-RPC 2.0 Model Context Protocol tool server with schema validation and isolated runtime errors.
+- [MCP tool server](mcp-server/README.md): build a stateless 2026-07-28 MCP tool server: no initialize handshake, schema validation, error isolation, and the input_required round-trip.
 - [Async batch fetcher](async-fetcher/README.md): control concurrency with Semaphore, back off and retry transient HTTP failures, and collect structured results.
 - [Resilient batch pipeline](structured-pipeline/README.md): process semi-structured data safely with batch error isolation and field validation using the standard library.