Merge pull request #341 from GeWuYou/feat/cqrs-optimization

Feat/cqrs optimization

.agents/skills/gframework-batch-boot/SKILL.md

@@ -12,6 +12,10 @@ batches until a clear stop condition is met.
 
 Treat `AGENTS.md` as the source of truth. This skill extends `gframework-boot`; it does not replace it.
 
+Context budget is a first-class stop signal. Do not keep batching merely because a file-count threshold still has
+headroom if the active conversation, loaded repo artifacts, validation output, and pending recovery updates suggest the
+agent is approaching its safe working-context limit.
+
 ## Startup Workflow
 
 1. Execute the normal `gframework-boot` startup sequence first:
@@ -28,6 +32,11 @@ Treat `AGENTS.md` as the source of truth. This skill extends `gframework-boot`;
    - repeated test refactor pattern
    - module-by-module documentation refresh
    - other repetitive multi-file cleanup
+4. Before the first implementation batch, estimate whether the current task is likely to stay below roughly 80% of the
+   agent's safe working-context budget through one more full batch cycle:
+   - include already loaded `AGENTS.md`, skills, `ai-plan` files, recent command output, active diffs, and expected validation output
+   - if another batch would probably push the conversation near the limit, plan to stop after the current batch even if
+     branch-size thresholds still have room
 
 ## Baseline Selection
 
@@ -67,8 +76,15 @@ For shorthand numeric thresholds, use a fixed default baseline:
 
 Choose one primary stop condition before the first batch and restate it to the user.
 
+When the user does not explicitly override the priority order, use:
+
+1. context-budget safety
+2. semantic batch boundary / reviewability
+3. the user-requested local metric such as files, lines, warnings, or time
+
 Common stop conditions:
 
+- the next batch would likely push the agent above roughly 80% of its safe working-context budget
 - branch diff vs baseline approaches a file-count threshold
 - warnings-only build reaches a target count
 - a specific hotspot list is exhausted
@@ -76,6 +92,9 @@ Common stop conditions:
 
 If multiple stop conditions exist, rank them and treat one as primary.
 
+Treat file-count or line-count thresholds as coarse repository-scope signals, not as a proxy for AI context health.
+When they disagree with context-budget safety, context-budget safety wins.
+
 ## Shorthand Stop-Condition Syntax
 
 `gframework-batch-boot` may be invoked with shorthand numeric thresholds when the user clearly wants a branch-size stop
@@ -108,6 +127,7 @@ When shorthand is used:
    - current branch and active topic
    - selected baseline
    - current stop-condition metric
+   - current context-budget posture and whether one more batch is safe
    - next candidate slices
 2. Keep the critical path local.
 3. Delegate only bounded slices with explicit ownership:
@@ -128,6 +148,7 @@ When shorthand is used:
    - integrate or verify the result
    - rerun the required validation
    - recompute the primary stop-condition metric
+   - reassess whether one more batch would likely push the agent near or beyond roughly 80% context usage
    - decide immediately whether to continue or stop
 7. Do not require the user to manually trigger every round unless:
    - the next slice is ambiguous
@@ -158,6 +179,7 @@ For multi-batch work, keep recovery artifacts current.
 
 Stop the loop when any of the following becomes true:
 
+- the next batch would likely push the agent near or beyond roughly 80% of its safe working-context budget
 - the primary stop condition has been reached or exceeded
 - the remaining slices are no longer low-risk
 - validation failures indicate the task is no longer repetitive
@@ -165,6 +187,7 @@ Stop the loop when any of the following becomes true:
 
 When stopping, report:
 
+- whether context budget was the deciding factor
 - which baseline was used
 - the exact metric value at stop time
 - completed batches

.agents/skills/gframework-boot/SKILL.md

@@ -36,14 +36,18 @@ Treat `AGENTS.md` as the source of truth. Use this skill to enforce a startup se
    - `simple`: one concern, one file or module, no parallel discovery required
    - `medium`: a small number of modules, some read-only exploration helpful, critical path still easy to keep local
    - `complex`: cross-module design, migration, large refactor, or work likely to exceed one context window
-11. Apply the delegation policy from `AGENTS.md`:
+11. Estimate the current context-budget posture before substantive execution:
+   - account for loaded startup artifacts, active `ai-plan` files, visible diffs, open validation output, and likely next-step output volume
+   - if the task already appears near roughly 80% of a safe working-context budget, prefer closing the current batch,
+     refreshing recovery artifacts, and stopping at the next natural semantic boundary instead of starting a fresh broad slice
+12. Apply the delegation policy from `AGENTS.md`:
    - Keep the critical path local
    - Use `explorer` with `gpt-5.1-codex-mini` for narrow read-only questions, tracing, inventory, and comparisons
    - Use `worker` with `gpt-5.4` only for bounded implementation tasks with explicit ownership
    - Do not delegate purely for ceremony; delegate only when it materially shortens the task or controls context growth
-12. Before editing files, tell the user what you read, how you classified the task, whether subagents will be used,
+13. Before editing files, tell the user what you read, how you classified the task, whether subagents will be used,
     and the first implementation step.
-13. Proceed with execution, validation, and documentation updates required by `AGENTS.md`.
+14. Proceed with execution, validation, and documentation updates required by `AGENTS.md`.
 
 ## Task Tracking
 
@@ -69,6 +73,8 @@ For multi-step, cross-module, or interruption-prone work, maintain the repositor
   first, then search the mapped active topics before scanning the broader public area.
 - If the current branch and the mapped active topics describe the same feature area, prefer resuming those topics first.
 - If the repository state suggests in-flight work but no recovery document matches, reconstruct the safest next step from code, tests, and Git state before asking the user for clarification.
+- If the current turn already carries heavy recovery context, broad diffs, or long validation output, prefer a
+  recovery-point update and a clean stop over starting another large slice just because the code task itself remains open.
 
 ## Example Triggers
 

GFramework.Core/Ioc/MicrosoftDiContainer.cs

@@ -185,6 +185,12 @@ public class MicrosoftDiContainer(IServiceCollection? serviceCollection = null)
     /// </summary>
     private IServiceProvider? _provider;
 
+    /// <summary>
+    ///     冻结后可复用的服务类型可见性索引。
+    ///     容器冻结后注册集合不再变化，因此 <see cref="HasRegistration(Type)" /> 可以安全复用该索引。
+    /// </summary>
+    private FrozenServiceTypeIndex? _frozenServiceTypeIndex;
+
     /// <summary>
     /// 容器冻结状态标志，true表示容器已冻结不可修改
     /// </summary>
@@ -1044,6 +1050,11 @@ public class MicrosoftDiContainer(IServiceCollection? serviceCollection = null)
         EnterReadLockOrThrowDisposed();
         try
         {
+            if (_frozenServiceTypeIndex is not null)
+            {
+                return _frozenServiceTypeIndex.Contains(type);
+            }
+
             return HasRegistrationCore(type);
         }
         finally
@@ -1139,6 +1150,7 @@ public class MicrosoftDiContainer(IServiceCollection? serviceCollection = null)
             GetServicesUnsafe.Clear();
             _registeredInstances.Clear();
             _provider = null;
+            _frozenServiceTypeIndex = null;
             _frozen = false;
             _logger.Info("Container cleared");
         }
@@ -1166,6 +1178,7 @@ public class MicrosoftDiContainer(IServiceCollection? serviceCollection = null)
             }
 
             _provider = GetServicesUnsafe.BuildServiceProvider();
+            _frozenServiceTypeIndex = FrozenServiceTypeIndex.Create(GetServicesUnsafe);
             _frozen = true;
             _logger.Info("IOC Container frozen - ServiceProvider built");
         }
@@ -1175,6 +1188,59 @@ public class MicrosoftDiContainer(IServiceCollection? serviceCollection = null)
         }
     }
 
+    /// <summary>
+    ///     保存冻结后按服务键可见的精确服务类型与开放泛型定义集合。
+    /// </summary>
+    /// <remarks>
+    ///     该索引只回答“按当前服务键语义是否可见”，因此与 <see cref="Get(Type)" /> /
+    ///     <see cref="GetAll(Type)" /> 一样不会退化为更宽松的可赋值匹配。
+    /// </remarks>
+    private sealed class FrozenServiceTypeIndex(HashSet<Type> exactServiceTypes, HashSet<Type> openGenericServiceTypes)
+    {
+        private readonly HashSet<Type> _exactServiceTypes = exactServiceTypes;
+        private readonly HashSet<Type> _openGenericServiceTypes = openGenericServiceTypes;
+
+        /// <summary>
+        ///     基于冻结时最终确定的服务描述符集合创建索引。
+        /// </summary>
+        /// <param name="descriptors">冻结时的服务描述符序列。</param>
+        /// <returns>供存在性判断热路径复用的服务键索引。</returns>
+        public static FrozenServiceTypeIndex Create(IEnumerable<ServiceDescriptor> descriptors)
+        {
+            ArgumentNullException.ThrowIfNull(descriptors);
+
+            var exactServiceTypes = new HashSet<Type>();
+            var openGenericServiceTypes = new HashSet<Type>();
+
+            foreach (var descriptor in descriptors)
+            {
+                var serviceType = descriptor.ServiceType;
+                exactServiceTypes.Add(serviceType);
+
+                if (serviceType.IsGenericTypeDefinition)
+                {
+                    openGenericServiceTypes.Add(serviceType);
+                }
+            }
+
+            return new FrozenServiceTypeIndex(exactServiceTypes, openGenericServiceTypes);
+        }
+
+        /// <summary>
+        ///     判断当前索引是否声明了目标服务键。
+        /// </summary>
+        /// <param name="requestedType">要检查的服务类型。</param>
+        /// <returns>命中精确服务键或可闭合的开放泛型服务键时返回 <see langword="true" />。</returns>
+        public bool Contains(Type requestedType)
+        {
+            ArgumentNullException.ThrowIfNull(requestedType);
+
+            return _exactServiceTypes.Contains(requestedType) ||
+                   requestedType.IsConstructedGenericType &&
+                   _openGenericServiceTypes.Contains(requestedType.GetGenericTypeDefinition());
+        }
+    }
+
     /// <summary>
     ///     获取底层的服务集合
     ///     提供对内部IServiceCollection的访问权限，用于高级配置和自定义操作
@@ -1250,6 +1316,7 @@ public class MicrosoftDiContainer(IServiceCollection? serviceCollection = null)
             _disposed = true;
             (_provider as IDisposable)?.Dispose();
             _provider = null;
+            _frozenServiceTypeIndex = null;
             GetServicesUnsafe.Clear();
             _registeredInstances.Clear();
             _frozen = false;

GFramework.Cqrs.Benchmarks/Messaging/BenchmarkHostFactory.cs

@@ -3,10 +3,13 @@
 
 using System;
 using System.Linq;
+using GFramework.Core.Abstractions.Logging;
 using GFramework.Core.Ioc;
 using GFramework.Cqrs.Abstractions.Cqrs;
+using GFramework.Cqrs.Internal;
 using MediatR;
 using Microsoft.Extensions.DependencyInjection;
+using LegacyICqrsRuntime = GFramework.Core.Abstractions.Cqrs.ICqrsRuntime;
 
 namespace GFramework.Cqrs.Benchmarks.Messaging;
 
@@ -31,11 +34,91 @@ internal static class BenchmarkHostFactory
         ArgumentNullException.ThrowIfNull(configure);
 
         var container = new MicrosoftDiContainer();
+        RegisterCqrsInfrastructure(container);
         configure(container);
         container.Freeze();
         return container;
     }
 
+    /// <summary>
+    ///     为 benchmark 宿主补齐默认 CQRS runtime seam，确保它既能手工注册 handler，也能走真实的程序集注册入口。
+    /// </summary>
+    /// <param name="container">当前 benchmark 拥有的 GFramework 容器。</param>
+    /// <remarks>
+    ///     `RegisterCqrsHandlersFromAssembly(...)` 依赖预先可见的 runtime / registrar / registration service 实例绑定。
+    ///     benchmark 宿主直接使用裸 <see cref="MicrosoftDiContainer" />，因此需要在配置阶段先补齐这组基础设施，
+    ///     避免各个 benchmark 用例各自复制同一段前置接线逻辑。
+    /// </remarks>
+    private static void RegisterCqrsInfrastructure(MicrosoftDiContainer container)
+    {
+        ArgumentNullException.ThrowIfNull(container);
+
+        if (container.Get<ICqrsRuntime>() is null)
+        {
+            var runtimeLogger = LoggerFactoryResolver.Provider.CreateLogger("CqrsDispatcher");
+            var notificationPublisher = container.Get<GFramework.Cqrs.Notification.INotificationPublisher>();
+            var runtime = GFramework.Cqrs.CqrsRuntimeFactory.CreateRuntime(container, runtimeLogger, notificationPublisher);
+            container.Register(runtime);
+            RegisterLegacyRuntimeAlias(container, runtime);
+        }
+        else if (container.Get<LegacyICqrsRuntime>() is null)
+        {
+            RegisterLegacyRuntimeAlias(container, container.GetRequired<ICqrsRuntime>());
+        }
+
+        if (container.Get<ICqrsHandlerRegistrar>() is null)
+        {
+            var registrarLogger = LoggerFactoryResolver.Provider.CreateLogger("DefaultCqrsHandlerRegistrar");
+            var registrar = GFramework.Cqrs.CqrsRuntimeFactory.CreateHandlerRegistrar(container, registrarLogger);
+            container.Register<ICqrsHandlerRegistrar>(registrar);
+        }
+
+        if (container.Get<ICqrsRegistrationService>() is null)
+        {
+            var registrationLogger = LoggerFactoryResolver.Provider.CreateLogger("DefaultCqrsRegistrationService");
+            var registrar = container.GetRequired<ICqrsHandlerRegistrar>();
+            var registrationService = GFramework.Cqrs.CqrsRuntimeFactory.CreateRegistrationService(registrar, registrationLogger);
+            container.Register<ICqrsRegistrationService>(registrationService);
+        }
+    }
+
+    /// <summary>
+    ///     只激活当前 benchmark 场景明确拥有的 generated registry，避免同一程序集里的其他 benchmark registry
+    ///     扩大冻结后服务索引与 dispatcher descriptor 基线。
+    /// </summary>
+    /// <typeparam name="TRegistry">当前 benchmark 需要接入的 generated registry 类型。</typeparam>
+    /// <param name="container">承载 generated registry 注册结果的 GFramework benchmark 容器。</param>
+    internal static void RegisterGeneratedBenchmarkRegistry<TRegistry>(MicrosoftDiContainer container)
+        where TRegistry : class, GFramework.Cqrs.ICqrsHandlerRegistry
+    {
+        ArgumentNullException.ThrowIfNull(container);
+
+        var registrarLogger = LoggerFactoryResolver.Provider.CreateLogger("DefaultCqrsHandlerRegistrar");
+        CqrsHandlerRegistrar.RegisterGeneratedRegistry(container, typeof(TRegistry), registrarLogger);
+    }
+
+    /// <summary>
+    ///     为旧命名空间下的 CQRS runtime 契约注册兼容别名。
+    /// </summary>
+    /// <param name="container">承载 runtime 别名的 benchmark 容器。</param>
+    /// <param name="runtime">当前正式 CQRS runtime 实例。</param>
+    /// <exception cref="InvalidOperationException">
+    ///     <paramref name="runtime" /> 未同时实现 legacy CQRS runtime 契约。
+    /// </exception>
+    private static void RegisterLegacyRuntimeAlias(MicrosoftDiContainer container, ICqrsRuntime runtime)
+    {
+        ArgumentNullException.ThrowIfNull(container);
+        ArgumentNullException.ThrowIfNull(runtime);
+
+        if (runtime is not LegacyICqrsRuntime legacyRuntime)
+        {
+            throw new InvalidOperationException(
+                $"The registered {typeof(ICqrsRuntime).FullName} must also implement {typeof(LegacyICqrsRuntime).FullName}. Actual runtime type: {runtime.GetType().FullName}.");
+        }
+
+        container.Register<LegacyICqrsRuntime>(legacyRuntime);
+    }
+
     /// <summary>
     ///     创建只承载当前 benchmark handler 集合的最小 MediatR 宿主。
     /// </summary>

GFramework.Cqrs.Benchmarks/Messaging/GeneratedDefaultRequestBenchmarkRegistry.cs

@@ -0,0 +1,100 @@
+// Copyright (c) 2025-2026 GeWuYou
+// SPDX-License-Identifier: Apache-2.0
+
+using System;
+using System.Collections.Generic;
+using System.Reflection;
+using System.Threading;
+using System.Threading.Tasks;
+using GFramework.Core.Abstractions.Logging;
+using GFramework.Cqrs.Abstractions.Cqrs;
+using Microsoft.Extensions.DependencyInjection;
+
+[assembly: GFramework.Cqrs.CqrsHandlerRegistryAttribute(
+    typeof(GFramework.Cqrs.Benchmarks.Messaging.GeneratedDefaultRequestBenchmarkRegistry))]
+
+namespace GFramework.Cqrs.Benchmarks.Messaging;
+
+/// <summary>
+///     为默认 request steady-state benchmark 提供 hand-written generated registry，
+///     以便验证“默认宿主吸收 generated request invoker provider”后的热路径收益。
+/// </summary>
+public sealed class GeneratedDefaultRequestBenchmarkRegistry :
+    GFramework.Cqrs.ICqrsHandlerRegistry,
+    GFramework.Cqrs.ICqrsRequestInvokerProvider,
+    GFramework.Cqrs.IEnumeratesCqrsRequestInvokerDescriptors
+{
+    private static readonly GFramework.Cqrs.CqrsRequestInvokerDescriptor Descriptor =
+        new(
+            typeof(IRequestHandler<
+                RequestBenchmarks.BenchmarkRequest,
+                RequestBenchmarks.BenchmarkResponse>),
+            typeof(GeneratedDefaultRequestBenchmarkRegistry).GetMethod(
+                nameof(InvokeBenchmarkRequestHandler),
+                BindingFlags.Public | BindingFlags.Static)
+            ?? throw new InvalidOperationException("Missing generated default request benchmark method."));
+
+    private static readonly IReadOnlyList<GFramework.Cqrs.CqrsRequestInvokerDescriptorEntry> Descriptors =
+    [
+        new GFramework.Cqrs.CqrsRequestInvokerDescriptorEntry(
+            typeof(RequestBenchmarks.BenchmarkRequest),
+            typeof(RequestBenchmarks.BenchmarkResponse),
+            Descriptor)
+    ];
+
+    /// <summary>
+    ///     把默认 request benchmark handler 注册为单例，保持与原先 steady-state 宿主一致的生命周期语义。
+    /// </summary>
+    public void Register(IServiceCollection services, ILogger logger)
+    {
+        ArgumentNullException.ThrowIfNull(services);
+        ArgumentNullException.ThrowIfNull(logger);
+
+        services.AddSingleton(
+            typeof(IRequestHandler<RequestBenchmarks.BenchmarkRequest, RequestBenchmarks.BenchmarkResponse>),
+            typeof(RequestBenchmarks.BenchmarkRequestHandler));
+        logger.Debug("Registered generated default request benchmark handler.");
+    }
+
+    /// <summary>
+    ///     返回当前 provider 暴露的全部 generated request invoker 描述符。
+    /// </summary>
+    public IReadOnlyList<GFramework.Cqrs.CqrsRequestInvokerDescriptorEntry> GetDescriptors()
+    {
+        return Descriptors;
+    }
+
+    /// <summary>
+    ///     为目标请求/响应类型对返回 generated request invoker 描述符。
+    /// </summary>
+    public bool TryGetDescriptor(
+        Type requestType,
+        Type responseType,
+        out GFramework.Cqrs.CqrsRequestInvokerDescriptor? descriptor)
+    {
+        if (requestType == typeof(RequestBenchmarks.BenchmarkRequest) &&
+            responseType == typeof(RequestBenchmarks.BenchmarkResponse))
+        {
+            descriptor = Descriptor;
+            return true;
+        }
+
+        descriptor = null;
+        return false;
+    }
+
+    /// <summary>
+    ///     模拟 generated invoker provider 为默认 request benchmark 产出的开放静态调用入口。
+    /// </summary>
+    public static ValueTask<RequestBenchmarks.BenchmarkResponse> InvokeBenchmarkRequestHandler(
+        object handler,
+        object request,
+        CancellationToken cancellationToken)
+    {
+        var typedHandler = (IRequestHandler<
+            RequestBenchmarks.BenchmarkRequest,
+            RequestBenchmarks.BenchmarkResponse>)handler;
+        var typedRequest = (RequestBenchmarks.BenchmarkRequest)request;
+        return typedHandler.Handle(typedRequest, cancellationToken);
+    }
+}

GFramework.Cqrs.Benchmarks/Messaging/GeneratedDefaultStreamingBenchmarkRegistry.cs

@@ -0,0 +1,96 @@
+// Copyright (c) 2025-2026 GeWuYou
+// SPDX-License-Identifier: Apache-2.0
+
+using System;
+using System.Collections.Generic;
+using System.Reflection;
+using System.Threading;
+using GFramework.Core.Abstractions.Logging;
+using GFramework.Cqrs.Abstractions.Cqrs;
+using Microsoft.Extensions.DependencyInjection;
+
+namespace GFramework.Cqrs.Benchmarks.Messaging;
+
+/// <summary>
+///     为默认 stream steady-state benchmark 提供 hand-written generated registry，
+///     以便验证“默认 stream 宿主吸收 generated stream invoker provider”后的完整枚举收益。
+/// </summary>
+public sealed class GeneratedDefaultStreamingBenchmarkRegistry :
+    GFramework.Cqrs.ICqrsHandlerRegistry,
+    GFramework.Cqrs.ICqrsStreamInvokerProvider,
+    GFramework.Cqrs.IEnumeratesCqrsStreamInvokerDescriptors
+{
+    private static readonly GFramework.Cqrs.CqrsStreamInvokerDescriptor Descriptor =
+        new(
+            typeof(IStreamRequestHandler<
+                StreamingBenchmarks.BenchmarkStreamRequest,
+                StreamingBenchmarks.BenchmarkResponse>),
+            typeof(GeneratedDefaultStreamingBenchmarkRegistry).GetMethod(
+                nameof(InvokeBenchmarkStreamHandler),
+                BindingFlags.Public | BindingFlags.Static)
+            ?? throw new InvalidOperationException("Missing generated default streaming benchmark method."));
+
+    private static readonly IReadOnlyList<GFramework.Cqrs.CqrsStreamInvokerDescriptorEntry> Descriptors =
+    [
+        new GFramework.Cqrs.CqrsStreamInvokerDescriptorEntry(
+            typeof(StreamingBenchmarks.BenchmarkStreamRequest),
+            typeof(StreamingBenchmarks.BenchmarkResponse),
+            Descriptor)
+    ];
+
+    /// <summary>
+    ///     把默认 stream benchmark handler 注册为单例，保持与原先 steady-state 宿主一致的生命周期语义。
+    /// </summary>
+    public void Register(IServiceCollection services, ILogger logger)
+    {
+        ArgumentNullException.ThrowIfNull(services);
+        ArgumentNullException.ThrowIfNull(logger);
+
+        services.AddSingleton(
+            typeof(IStreamRequestHandler<StreamingBenchmarks.BenchmarkStreamRequest, StreamingBenchmarks.BenchmarkResponse>),
+            typeof(StreamingBenchmarks.BenchmarkStreamHandler));
+        logger.Debug("Registered generated default streaming benchmark handler.");
+    }
+
+    /// <summary>
+    ///     返回当前 provider 暴露的全部 generated stream invoker 描述符。
+    /// </summary>
+    public IReadOnlyList<GFramework.Cqrs.CqrsStreamInvokerDescriptorEntry> GetDescriptors()
+    {
+        return Descriptors;
+    }
+
+    /// <summary>
+    ///     为目标流式请求/响应类型对返回 generated stream invoker 描述符。
+    /// </summary>
+    public bool TryGetDescriptor(
+        Type requestType,
+        Type responseType,
+        out GFramework.Cqrs.CqrsStreamInvokerDescriptor? descriptor)
+    {
+        if (requestType == typeof(StreamingBenchmarks.BenchmarkStreamRequest) &&
+            responseType == typeof(StreamingBenchmarks.BenchmarkResponse))
+        {
+            descriptor = Descriptor;
+            return true;
+        }
+
+        descriptor = null;
+        return false;
+    }
+
+    /// <summary>
+    ///     模拟 generated stream invoker provider 为默认 stream benchmark 产出的开放静态调用入口。
+    /// </summary>
+    public static object InvokeBenchmarkStreamHandler(
+        object handler,
+        object request,
+        CancellationToken cancellationToken)
+    {
+        var typedHandler = (IStreamRequestHandler<
+            StreamingBenchmarks.BenchmarkStreamRequest,
+            StreamingBenchmarks.BenchmarkResponse>)handler;
+        var typedRequest = (StreamingBenchmarks.BenchmarkStreamRequest)request;
+        return typedHandler.Handle(typedRequest, cancellationToken);
+    }
+}

GFramework.Cqrs.Benchmarks/Messaging/GeneratedRequestPipelineBenchmarkRegistry.cs

@@ -0,0 +1,100 @@
+// Copyright (c) 2025-2026 GeWuYou
+// SPDX-License-Identifier: Apache-2.0
+
+using System;
+using System.Collections.Generic;
+using System.Reflection;
+using System.Threading;
+using System.Threading.Tasks;
+using GFramework.Core.Abstractions.Logging;
+using GFramework.Cqrs.Abstractions.Cqrs;
+using Microsoft.Extensions.DependencyInjection;
+
+[assembly: GFramework.Cqrs.CqrsHandlerRegistryAttribute(
+    typeof(GFramework.Cqrs.Benchmarks.Messaging.GeneratedRequestPipelineBenchmarkRegistry))]
+
+namespace GFramework.Cqrs.Benchmarks.Messaging;
+
+/// <summary>
+///     为 request pipeline benchmark 提供 handwritten generated registry，
+///     让默认 pipeline 宿主也能走真实的 generated request invoker provider 接线路径。
+/// </summary>
+public sealed class GeneratedRequestPipelineBenchmarkRegistry :
+    GFramework.Cqrs.ICqrsHandlerRegistry,
+    GFramework.Cqrs.ICqrsRequestInvokerProvider,
+    GFramework.Cqrs.IEnumeratesCqrsRequestInvokerDescriptors
+{
+    private static readonly GFramework.Cqrs.CqrsRequestInvokerDescriptor Descriptor =
+        new(
+            typeof(IRequestHandler<
+                RequestPipelineBenchmarks.BenchmarkRequest,
+                RequestPipelineBenchmarks.BenchmarkResponse>),
+            typeof(GeneratedRequestPipelineBenchmarkRegistry).GetMethod(
+                nameof(InvokeBenchmarkRequestHandler),
+                BindingFlags.Public | BindingFlags.Static)
+            ?? throw new InvalidOperationException("Missing generated request pipeline benchmark method."));
+
+    private static readonly IReadOnlyList<GFramework.Cqrs.CqrsRequestInvokerDescriptorEntry> Descriptors =
+    [
+        new GFramework.Cqrs.CqrsRequestInvokerDescriptorEntry(
+            typeof(RequestPipelineBenchmarks.BenchmarkRequest),
+            typeof(RequestPipelineBenchmarks.BenchmarkResponse),
+            Descriptor)
+    ];
+
+    /// <summary>
+    ///     将 request pipeline benchmark handler 注册为单例，保持与当前矩阵宿主一致的生命周期语义。
+    /// </summary>
+    public void Register(IServiceCollection services, ILogger logger)
+    {
+        ArgumentNullException.ThrowIfNull(services);
+        ArgumentNullException.ThrowIfNull(logger);
+
+        services.AddSingleton(
+            typeof(IRequestHandler<RequestPipelineBenchmarks.BenchmarkRequest, RequestPipelineBenchmarks.BenchmarkResponse>),
+            typeof(RequestPipelineBenchmarks.BenchmarkRequestHandler));
+        logger.Debug("Registered generated request pipeline benchmark handler.");
+    }
+
+    /// <summary>
+    ///     返回当前 provider 暴露的全部 generated request invoker 描述符。
+    /// </summary>
+    public IReadOnlyList<GFramework.Cqrs.CqrsRequestInvokerDescriptorEntry> GetDescriptors()
+    {
+        return Descriptors;
+    }
+
+    /// <summary>
+    ///     为目标请求/响应类型对返回 generated request invoker 描述符。
+    /// </summary>
+    public bool TryGetDescriptor(
+        Type requestType,
+        Type responseType,
+        out GFramework.Cqrs.CqrsRequestInvokerDescriptor? descriptor)
+    {
+        if (requestType == typeof(RequestPipelineBenchmarks.BenchmarkRequest) &&
+            responseType == typeof(RequestPipelineBenchmarks.BenchmarkResponse))
+        {
+            descriptor = Descriptor;
+            return true;
+        }
+
+        descriptor = null;
+        return false;
+    }
+
+    /// <summary>
+    ///     模拟 generated invoker provider 为 request pipeline benchmark 产出的开放静态调用入口。
+    /// </summary>
+    public static ValueTask<RequestPipelineBenchmarks.BenchmarkResponse> InvokeBenchmarkRequestHandler(
+        object handler,
+        object request,
+        CancellationToken cancellationToken)
+    {
+        var typedHandler = (IRequestHandler<
+            RequestPipelineBenchmarks.BenchmarkRequest,
+            RequestPipelineBenchmarks.BenchmarkResponse>)handler;
+        var typedRequest = (RequestPipelineBenchmarks.BenchmarkRequest)request;
+        return typedHandler.Handle(typedRequest, cancellationToken);
+    }
+}

GFramework.Cqrs.Benchmarks/Messaging/GeneratedStreamLifetimeBenchmarkRegistry.cs

@@ -0,0 +1,110 @@
+// Copyright (c) 2025-2026 GeWuYou
+// SPDX-License-Identifier: Apache-2.0
+
+using System;
+using System.Collections.Generic;
+using System.Reflection;
+using System.Threading;
+using GFramework.Core.Abstractions.Logging;
+using GFramework.Cqrs.Abstractions.Cqrs;
+using Microsoft.Extensions.DependencyInjection;
+
+[assembly: GFramework.Cqrs.CqrsHandlerRegistryAttribute(
+    typeof(GFramework.Cqrs.Benchmarks.Messaging.GeneratedStreamLifetimeBenchmarkRegistry))]
+
+namespace GFramework.Cqrs.Benchmarks.Messaging;
+
+/// <summary>
+///     为 stream 生命周期矩阵 benchmark 提供 hand-written generated registry，
+///     以便在默认 generated-provider 宿主路径上比较不同 handler 生命周期的完整枚举成本。
+/// </summary>
+public sealed class GeneratedStreamLifetimeBenchmarkRegistry :
+    GFramework.Cqrs.ICqrsHandlerRegistry,
+    GFramework.Cqrs.ICqrsStreamInvokerProvider,
+    GFramework.Cqrs.IEnumeratesCqrsStreamInvokerDescriptors
+{
+    private static readonly GFramework.Cqrs.CqrsStreamInvokerDescriptor Descriptor =
+        new(
+            typeof(IStreamRequestHandler<
+                StreamLifetimeBenchmarks.BenchmarkStreamRequest,
+                StreamLifetimeBenchmarks.BenchmarkResponse>),
+            typeof(GeneratedStreamLifetimeBenchmarkRegistry).GetMethod(
+                nameof(InvokeBenchmarkStreamHandler),
+                BindingFlags.Public | BindingFlags.Static)
+            ?? throw new InvalidOperationException("Missing generated stream lifetime benchmark method."));
+
+    private static readonly IReadOnlyList<GFramework.Cqrs.CqrsStreamInvokerDescriptorEntry> Descriptors =
+    [
+        new GFramework.Cqrs.CqrsStreamInvokerDescriptorEntry(
+            typeof(StreamLifetimeBenchmarks.BenchmarkStreamRequest),
+            typeof(StreamLifetimeBenchmarks.BenchmarkResponse),
+            Descriptor)
+    ];
+
+    /// <summary>
+    ///     参与程序集注册入口，但不在这里直接写入 handler 生命周期。
+    /// </summary>
+    /// <param name="services">当前 generated registry 拥有的服务集合。</param>
+    /// <param name="logger">用于记录 generated registry 注册行为的日志器。</param>
+    /// <remarks>
+    ///     生命周期矩阵需要让 benchmark 主体显式控制 `Singleton / Transient` 变量。
+    ///     因此 registry 只负责暴露 generated descriptor，不在这里抢先注册 handler，避免把默认单例注册混入比较结果。
+    /// </remarks>
+    public void Register(IServiceCollection services, ILogger logger)
+    {
+        ArgumentNullException.ThrowIfNull(services);
+        ArgumentNullException.ThrowIfNull(logger);
+
+        logger.Debug("Registered generated stream lifetime benchmark descriptors.");
+    }
+
+    /// <summary>
+    ///     返回当前 provider 暴露的全部 generated stream invoker 描述符。
+    /// </summary>
+    public IReadOnlyList<GFramework.Cqrs.CqrsStreamInvokerDescriptorEntry> GetDescriptors()
+    {
+        return Descriptors;
+    }
+
+    /// <summary>
+    ///     为目标流式请求/响应类型对返回 generated stream invoker 描述符。
+    /// </summary>
+    /// <param name="requestType">待匹配的请求类型。</param>
+    /// <param name="responseType">待匹配的响应类型。</param>
+    /// <param name="descriptor">命中时返回的 generated descriptor。</param>
+    /// <returns>命中当前 benchmark 的请求/响应类型对时返回 <see langword="true" />。</returns>
+    public bool TryGetDescriptor(
+        Type requestType,
+        Type responseType,
+        out GFramework.Cqrs.CqrsStreamInvokerDescriptor? descriptor)
+    {
+        if (requestType == typeof(StreamLifetimeBenchmarks.BenchmarkStreamRequest) &&
+            responseType == typeof(StreamLifetimeBenchmarks.BenchmarkResponse))
+        {
+            descriptor = Descriptor;
+            return true;
+        }
+
+        descriptor = null;
+        return false;
+    }
+
+    /// <summary>
+    ///     模拟 generated stream invoker provider 为生命周期矩阵 benchmark 产出的开放静态调用入口。
+    /// </summary>
+    /// <param name="handler">当前请求对应的 handler 实例。</param>
+    /// <param name="request">待分发的流式请求。</param>
+    /// <param name="cancellationToken">调用方传入的取消令牌。</param>
+    /// <returns>交给目标 stream handler 处理后的异步枚举。</returns>
+    public static object InvokeBenchmarkStreamHandler(
+        object handler,
+        object request,
+        CancellationToken cancellationToken)
+    {
+        var typedHandler = (IStreamRequestHandler<
+            StreamLifetimeBenchmarks.BenchmarkStreamRequest,
+            StreamLifetimeBenchmarks.BenchmarkResponse>)handler;
+        var typedRequest = (StreamLifetimeBenchmarks.BenchmarkStreamRequest)request;
+        return typedHandler.Handle(typedRequest, cancellationToken);
+    }
+}

GFramework.Cqrs.Benchmarks/Messaging/RequestBenchmarks.cs

@@ -61,12 +61,12 @@ public class RequestBenchmarks
             MinLevel = LogLevel.Fatal
         };
         Fixture.Setup("Request", handlerCount: 1, pipelineCount: 0);
+        BenchmarkDispatcherCacheHelper.ClearDispatcherCaches();
 
         _baselineHandler = new BenchmarkRequestHandler();
         _container = BenchmarkHostFactory.CreateFrozenGFrameworkContainer(container =>
         {
-            container.RegisterSingleton<GFramework.Cqrs.Abstractions.Cqrs.IRequestHandler<BenchmarkRequest, BenchmarkResponse>>(
-                _baselineHandler);
+            BenchmarkHostFactory.RegisterGeneratedBenchmarkRegistry<GeneratedDefaultRequestBenchmarkRegistry>(container);
         });
         _runtime = GFramework.Cqrs.CqrsRuntimeFactory.CreateRuntime(
             _container,
@@ -91,7 +91,14 @@ public class RequestBenchmarks
     [GlobalCleanup]
     public void Cleanup()
     {
-        BenchmarkCleanupHelper.DisposeAll(_container, _mediatrServiceProvider, _mediatorServiceProvider);
+        try
+        {
+            BenchmarkCleanupHelper.DisposeAll(_container, _mediatrServiceProvider, _mediatorServiceProvider);
+        }
+        finally
+        {
+            BenchmarkDispatcherCacheHelper.ClearDispatcherCaches();
+        }
     }
 
     /// <summary>

GFramework.Cqrs.Benchmarks/Messaging/RequestInvokerBenchmarks.cs

@@ -83,7 +83,7 @@ public class RequestInvokerBenchmarks
 
         _generatedContainer = BenchmarkHostFactory.CreateFrozenGFrameworkContainer(container =>
         {
-            container.RegisterCqrsHandlersFromAssembly(typeof(RequestInvokerBenchmarks).Assembly);
+            BenchmarkHostFactory.RegisterGeneratedBenchmarkRegistry<GeneratedRequestInvokerBenchmarkRegistry>(container);
         });
         _generatedRuntime = GFramework.Cqrs.CqrsRuntimeFactory.CreateRuntime(
             _generatedContainer,

GFramework.Cqrs.Benchmarks/Messaging/RequestPipelineBenchmarks.cs

@@ -69,8 +69,7 @@ public class RequestPipelineBenchmarks
         _baselineHandler = new BenchmarkRequestHandler();
         _container = BenchmarkHostFactory.CreateFrozenGFrameworkContainer(container =>
         {
-            container.RegisterSingleton<GFramework.Cqrs.Abstractions.Cqrs.IRequestHandler<BenchmarkRequest, BenchmarkResponse>>(
-                _baselineHandler);
+            BenchmarkHostFactory.RegisterGeneratedBenchmarkRegistry<GeneratedRequestPipelineBenchmarkRegistry>(container);
             RegisterGFrameworkPipelineBehaviors(container, PipelineCount);
         });
         _runtime = GFramework.Cqrs.CqrsRuntimeFactory.CreateRuntime(

GFramework.Cqrs.Benchmarks/Messaging/StreamInvokerBenchmarks.cs

@@ -83,7 +83,7 @@ public class StreamInvokerBenchmarks
 
         _generatedContainer = BenchmarkHostFactory.CreateFrozenGFrameworkContainer(container =>
         {
-            container.RegisterCqrsHandlersFromAssembly(typeof(StreamInvokerBenchmarks).Assembly);
+            BenchmarkHostFactory.RegisterGeneratedBenchmarkRegistry<GeneratedStreamInvokerBenchmarkRegistry>(container);
         });
         _generatedRuntime = GFramework.Cqrs.CqrsRuntimeFactory.CreateRuntime(
             _generatedContainer,

GFramework.Cqrs.Benchmarks/Messaging/StreamLifetimeBenchmarks.cs

@@ -0,0 +1,279 @@
+// Copyright (c) 2025-2026 GeWuYou
+// SPDX-License-Identifier: Apache-2.0
+
+using BenchmarkDotNet.Attributes;
+using BenchmarkDotNet.Columns;
+using BenchmarkDotNet.Configs;
+using BenchmarkDotNet.Diagnosers;
+using BenchmarkDotNet.Jobs;
+using BenchmarkDotNet.Order;
+using System;
+using System.Collections.Generic;
+using System.Threading;
+using System.Threading.Tasks;
+using GFramework.Core.Abstractions.Logging;
+using GFramework.Core.Ioc;
+using GFramework.Core.Logging;
+using GFramework.Cqrs.Abstractions.Cqrs;
+using MediatR;
+using Microsoft.Extensions.DependencyInjection;
+
+namespace GFramework.Cqrs.Benchmarks.Messaging;
+
+/// <summary>
+///     对比 stream 完整枚举在不同 handler 生命周期下的额外开销。
+/// </summary>
+/// <remarks>
+///     当前矩阵只覆盖 `Singleton` 与 `Transient`。
+///     `Scoped` 仍依赖真实的显式作用域边界；在当前“单根容器最小宿主”模型下直接加入 scoped 会把枚举宿主成本与生命周期成本混在一起，
+///     因此保持与 request 生命周期矩阵相同的边界，留待后续 scoped host 基线具备后再扩展。
+/// </remarks>
+[Config(typeof(Config))]
+public class StreamLifetimeBenchmarks
+{
+    private MicrosoftDiContainer _container = null!;
+    private ICqrsRuntime _runtime = null!;
+    private ServiceProvider _serviceProvider = null!;
+    private IMediator _mediatr = null!;
+    private BenchmarkStreamHandler _baselineHandler = null!;
+    private BenchmarkStreamRequest _request = null!;
+
+    /// <summary>
+    ///     控制当前 benchmark 使用的 handler 生命周期。
+    /// </summary>
+    [Params(HandlerLifetime.Singleton, HandlerLifetime.Transient)]
+    public HandlerLifetime Lifetime { get; set; }
+
+    /// <summary>
+    ///     可公平比较的 benchmark handler 生命周期集合。
+    /// </summary>
+    public enum HandlerLifetime
+    {
+        /// <summary>
+        ///     复用单个 handler 实例。
+        /// </summary>
+        Singleton,
+
+        /// <summary>
+        ///     每次建流都重新解析新的 handler 实例。
+        /// </summary>
+        Transient
+    }
+
+    /// <summary>
+    ///     配置 stream 生命周期 benchmark 的公共输出格式。
+    /// </summary>
+    private sealed class Config : ManualConfig
+    {
+        public Config()
+        {
+            AddJob(Job.Default);
+            AddColumnProvider(DefaultColumnProviders.Instance);
+            AddColumn(new CustomColumn("Scenario", static (_, _) => "StreamLifetime"));
+            AddDiagnoser(MemoryDiagnoser.Default);
+            WithOrderer(new DefaultOrderer(SummaryOrderPolicy.FastestToSlowest, MethodOrderPolicy.Declared));
+        }
+    }
+
+    /// <summary>
+    ///     构建当前生命周期下的 GFramework 与 MediatR stream 对照宿主。
+    /// </summary>
+    [GlobalSetup]
+    public void Setup()
+    {
+        LoggerFactoryResolver.Provider = new ConsoleLoggerFactoryProvider
+        {
+            MinLevel = LogLevel.Fatal
+        };
+        Fixture.Setup($"StreamLifetime/{Lifetime}", handlerCount: 1, pipelineCount: 0);
+        BenchmarkDispatcherCacheHelper.ClearDispatcherCaches();
+
+        _baselineHandler = new BenchmarkStreamHandler();
+        _request = new BenchmarkStreamRequest(Guid.NewGuid(), 3);
+
+        _container = BenchmarkHostFactory.CreateFrozenGFrameworkContainer(container =>
+        {
+            BenchmarkHostFactory.RegisterGeneratedBenchmarkRegistry<GeneratedStreamLifetimeBenchmarkRegistry>(container);
+            RegisterGFrameworkHandler(container, Lifetime);
+        });
+        // 容器内已提前保留默认 runtime 以支撑 generated registry 接线；
+        // 这里额外创建带生命周期后缀的 runtime，只是为了区分不同 benchmark 矩阵的 dispatcher 日志。
+        _runtime = GFramework.Cqrs.CqrsRuntimeFactory.CreateRuntime(
+            _container,
+            LoggerFactoryResolver.Provider.CreateLogger(nameof(StreamLifetimeBenchmarks) + "." + Lifetime));
+
+        _serviceProvider = BenchmarkHostFactory.CreateMediatRServiceProvider(
+            configure: null,
+            typeof(StreamLifetimeBenchmarks),
+            static candidateType => candidateType == typeof(BenchmarkStreamHandler),
+            ResolveMediatRLifetime(Lifetime));
+        _mediatr = _serviceProvider.GetRequiredService<IMediator>();
+    }
+
+    /// <summary>
+    ///     释放当前生命周期矩阵持有的 benchmark 宿主资源，并清理 dispatcher 缓存。
+    /// </summary>
+    [GlobalCleanup]
+    public void Cleanup()
+    {
+        try
+        {
+            BenchmarkCleanupHelper.DisposeAll(_container, _serviceProvider);
+        }
+        finally
+        {
+            BenchmarkDispatcherCacheHelper.ClearDispatcherCaches();
+        }
+    }
+
+    /// <summary>
+    ///     直接调用 handler 并完整枚举，作为不同生命周期矩阵下的 dispatch 额外开销 baseline。
+    /// </summary>
+    [Benchmark(Baseline = true)]
+    public async ValueTask Stream_Baseline()
+    {
+        await foreach (var response in _baselineHandler.Handle(_request, CancellationToken.None).ConfigureAwait(false))
+        {
+            _ = response;
+        }
+    }
+
+    /// <summary>
+    ///     通过 GFramework.CQRS runtime 创建并完整枚举 stream。
+    /// </summary>
+    [Benchmark]
+    public async ValueTask Stream_GFrameworkCqrs()
+    {
+        await foreach (var response in _runtime.CreateStream(BenchmarkContext.Instance, _request, CancellationToken.None)
+                           .ConfigureAwait(false))
+        {
+            _ = response;
+        }
+    }
+
+    /// <summary>
+    ///     通过 MediatR 创建并完整枚举 stream，作为外部对照。
+    /// </summary>
+    [Benchmark]
+    public async ValueTask Stream_MediatR()
+    {
+        await foreach (var response in _mediatr.CreateStream(_request, CancellationToken.None).ConfigureAwait(false))
+        {
+            _ = response;
+        }
+    }
+
+    /// <summary>
+    ///     按生命周期把 benchmark stream handler 注册到 GFramework 容器。
+    /// </summary>
+    /// <param name="container">当前 benchmark 拥有并负责释放的容器。</param>
+    /// <param name="lifetime">待比较的 handler 生命周期。</param>
+    /// <remarks>
+    ///     先通过 generated registry 提供静态 descriptor，再显式覆盖 handler 生命周期，
+    ///     可以把比较变量收敛到 handler 解析成本，而不是 descriptor 发现路径本身。
+    /// </remarks>
+    private static void RegisterGFrameworkHandler(MicrosoftDiContainer container, HandlerLifetime lifetime)
+    {
+        ArgumentNullException.ThrowIfNull(container);
+
+        switch (lifetime)
+        {
+            case HandlerLifetime.Singleton:
+                container.RegisterSingleton<
+                    GFramework.Cqrs.Abstractions.Cqrs.IStreamRequestHandler<BenchmarkStreamRequest, BenchmarkResponse>,
+                    BenchmarkStreamHandler>();
+                return;
+
+            case HandlerLifetime.Transient:
+                container.RegisterTransient<
+                    GFramework.Cqrs.Abstractions.Cqrs.IStreamRequestHandler<BenchmarkStreamRequest, BenchmarkResponse>,
+                    BenchmarkStreamHandler>();
+                return;
+
+            default:
+                throw new ArgumentOutOfRangeException(nameof(lifetime), lifetime, "Unsupported benchmark handler lifetime.");
+        }
+    }
+
+    /// <summary>
+    ///     将 benchmark 生命周期映射为 MediatR 组装所需的 <see cref="ServiceLifetime" />。
+    /// </summary>
+    /// <param name="lifetime">待比较的 handler 生命周期。</param>
+    /// <returns>当前生命周期对应的 MediatR 注册方式。</returns>
+    private static ServiceLifetime ResolveMediatRLifetime(HandlerLifetime lifetime)
+    {
+        return lifetime switch
+        {
+            HandlerLifetime.Singleton => ServiceLifetime.Singleton,
+            HandlerLifetime.Transient => ServiceLifetime.Transient,
+            _ => throw new ArgumentOutOfRangeException(nameof(lifetime), lifetime, "Unsupported benchmark handler lifetime.")
+        };
+    }
+
+    /// <summary>
+    ///     Benchmark stream request。
+    /// </summary>
+    /// <param name="Id">请求标识。</param>
+    /// <param name="ItemCount">返回元素数量。</param>
+    public sealed record BenchmarkStreamRequest(Guid Id, int ItemCount) :
+        GFramework.Cqrs.Abstractions.Cqrs.IStreamRequest<BenchmarkResponse>,
+        MediatR.IStreamRequest<BenchmarkResponse>;
+
+    /// <summary>
+    ///     Benchmark stream response。
+    /// </summary>
+    /// <param name="Id">响应标识。</param>
+    public sealed record BenchmarkResponse(Guid Id);
+
+    /// <summary>
+    ///     同时实现 GFramework.CQRS 与 MediatR 契约的最小 stream handler。
+    /// </summary>
+    public sealed class BenchmarkStreamHandler :
+        GFramework.Cqrs.Abstractions.Cqrs.IStreamRequestHandler<BenchmarkStreamRequest, BenchmarkResponse>,
+        MediatR.IStreamRequestHandler<BenchmarkStreamRequest, BenchmarkResponse>
+    {
+        /// <summary>
+        ///     处理 GFramework.CQRS stream request。
+        /// </summary>
+        /// <param name="request">当前 benchmark stream 请求。</param>
+        /// <param name="cancellationToken">用于中断异步枚举的取消令牌。</param>
+        /// <returns>完整枚举所需的低噪声异步响应序列。</returns>
+        public IAsyncEnumerable<BenchmarkResponse> Handle(
+            BenchmarkStreamRequest request,
+            CancellationToken cancellationToken)
+        {
+            return EnumerateAsync(request, cancellationToken);
+        }
+
+        /// <summary>
+        ///     处理 MediatR stream request。
+        /// </summary>
+        /// <param name="request">当前 benchmark stream 请求。</param>
+        /// <param name="cancellationToken">用于中断异步枚举的取消令牌。</param>
+        /// <returns>完整枚举所需的低噪声异步响应序列。</returns>
+        IAsyncEnumerable<BenchmarkResponse> MediatR.IStreamRequestHandler<BenchmarkStreamRequest, BenchmarkResponse>.Handle(
+            BenchmarkStreamRequest request,
+            CancellationToken cancellationToken)
+        {
+            return EnumerateAsync(request, cancellationToken);
+        }
+
+        /// <summary>
+        ///     为生命周期矩阵构造稳定、低噪声的异步响应序列。
+        /// </summary>
+        /// <param name="request">当前 benchmark 请求。</param>
+        /// <param name="cancellationToken">用于中断异步枚举的取消令牌。</param>
+        /// <returns>按固定元素数量返回的异步响应序列。</returns>
+        private static async IAsyncEnumerable<BenchmarkResponse> EnumerateAsync(
+            BenchmarkStreamRequest request,
+            [System.Runtime.CompilerServices.EnumeratorCancellation] CancellationToken cancellationToken)
+        {
+            for (var index = 0; index < request.ItemCount; index++)
+            {
+                cancellationToken.ThrowIfCancellationRequested();
+                yield return new BenchmarkResponse(request.Id);
+                await Task.CompletedTask.ConfigureAwait(false);
+            }
+        }
+    }
+}

GFramework.Cqrs.Benchmarks/Messaging/StreamingBenchmarks.cs

@@ -18,6 +18,9 @@ using GFramework.Cqrs.Abstractions.Cqrs;
 using MediatR;
 using Microsoft.Extensions.DependencyInjection;
 
+[assembly: GFramework.Cqrs.CqrsHandlerRegistryAttribute(
+    typeof(GFramework.Cqrs.Benchmarks.Messaging.GeneratedDefaultStreamingBenchmarkRegistry))]
+
 namespace GFramework.Cqrs.Benchmarks.Messaging;
 
 /// <summary>
@@ -59,12 +62,12 @@ public class StreamingBenchmarks
             MinLevel = LogLevel.Fatal
         };
         Fixture.Setup("StreamRequest", handlerCount: 1, pipelineCount: 0);
+        BenchmarkDispatcherCacheHelper.ClearDispatcherCaches();
 
         _baselineHandler = new BenchmarkStreamHandler();
         _container = BenchmarkHostFactory.CreateFrozenGFrameworkContainer(container =>
         {
-            container.RegisterSingleton<GFramework.Cqrs.Abstractions.Cqrs.IStreamRequestHandler<BenchmarkStreamRequest, BenchmarkResponse>>(
-                _baselineHandler);
+            BenchmarkHostFactory.RegisterGeneratedBenchmarkRegistry<GeneratedDefaultStreamingBenchmarkRegistry>(container);
         });
         _runtime = GFramework.Cqrs.CqrsRuntimeFactory.CreateRuntime(
             _container,
@@ -86,7 +89,14 @@ public class StreamingBenchmarks
     [GlobalCleanup]
     public void Cleanup()
     {
-        BenchmarkCleanupHelper.DisposeAll(_container, _serviceProvider);
+        try
+        {
+            BenchmarkCleanupHelper.DisposeAll(_container, _serviceProvider);
+        }
+        finally
+        {
+            BenchmarkDispatcherCacheHelper.ClearDispatcherCaches();
+        }
     }
 
     /// <summary>

GFramework.Cqrs.Benchmarks/README.md

@@ -15,11 +15,13 @@
 - `Messaging/Fixture.cs`
   - 运行前输出并校验场景配置
 - `Messaging/RequestBenchmarks.cs`
-  - direct handler、NuGet `Mediator` source-generated concrete path、`GFramework.Cqrs` runtime 与 `MediatR` 的 request steady-state dispatch 对比
+  - direct handler、NuGet `Mediator` source-generated concrete path、已接上 handwritten generated request invoker provider 的默认 `GFramework.Cqrs` runtime 与 `MediatR` 的 request steady-state dispatch 对比
 - `Messaging/RequestLifetimeBenchmarks.cs`
   - `Singleton / Transient` 两类 handler 生命周期下，direct handler、`GFramework.Cqrs` runtime 与 `MediatR` 的 request steady-state dispatch 对比
+- `Messaging/StreamLifetimeBenchmarks.cs`
+  - `Singleton / Transient` 两类 handler 生命周期下，direct handler、已接上 handwritten generated stream invoker provider 的 `GFramework.Cqrs` runtime 与 `MediatR` 的 stream 完整枚举对比
 - `Messaging/RequestPipelineBenchmarks.cs`
-  - `0 / 1 / 4` 个 pipeline 行为下，direct handler、`GFramework.Cqrs` runtime 与 `MediatR` 的 request steady-state dispatch 对比
+  - `0 / 1 / 4` 个 pipeline 行为下，direct handler、已接上 handwritten generated request invoker provider 的 `GFramework.Cqrs` runtime 与 `MediatR` 的 request steady-state dispatch 对比
 - `Messaging/RequestStartupBenchmarks.cs`
   - `Initialization` 与 `ColdStart` 两组 request startup 成本对比，补齐与 `Mediator` comparison benchmark 更接近的 startup 维度
 - `Messaging/RequestInvokerBenchmarks.cs`
@@ -29,7 +31,7 @@
 - `Messaging/NotificationBenchmarks.cs`
   - `GFramework.Cqrs` runtime 与 `MediatR` 的单处理器 notification publish 对比
 - `Messaging/StreamingBenchmarks.cs`
-  - direct handler、`GFramework.Cqrs` runtime 与 `MediatR` 的 stream request 完整枚举对比
+  - direct handler、已接上 handwritten generated stream invoker provider 的 `GFramework.Cqrs` runtime 与 `MediatR` 的 stream request 完整枚举对比
 
 ## 最小使用方式
 
@@ -51,6 +53,5 @@ dotnet run --project GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.cspro
 
 - request / stream 的真实 source-generator 产物与 handwritten generated provider 对照
 - `Mediator` 的 transient / scoped compile-time lifetime 矩阵对照
-- stream handler 生命周期矩阵
 - 带真实显式作用域边界的 scoped host 对照
 - generated invoker provider 与纯反射 dispatch / 建流对比继续扩展到更多场景

GFramework.Cqrs.Tests/Cqrs/CqrsDispatcherContextValidationTests.cs

@@ -43,6 +43,63 @@ internal sealed class CqrsDispatcherContextValidationTests
             Throws.InvalidOperationException.With.Message.Contains("does not implement IArchitectureContext"));
     }
 
+    /// <summary>
+    ///     验证 request 上下文校验失败时，<see cref="GFramework.Cqrs.Abstractions.Cqrs.ICqrsRuntime.SendAsync{TResponse}" />
+    ///     不会在调用点同步抛出，而是返回一个 faulted <see cref="ValueTask{TResult}" /> 保持既有异步失败语义。
+    /// </summary>
+    [Test]
+    public void SendAsync_Should_Return_Faulted_ValueTask_When_Context_Preparation_Fails()
+    {
+        var runtime = CreateRuntime(
+            container =>
+            {
+                container
+                    .Setup(currentContainer => currentContainer.Get(typeof(IRequestHandler<ContextAwareRequest, int>)))
+                    .Returns(new ContextAwareRequestHandler());
+                container
+                    .Setup(currentContainer => currentContainer.HasRegistration(typeof(IPipelineBehavior<ContextAwareRequest, int>)))
+                    .Returns(false);
+            });
+
+        ValueTask<int> dispatch = default;
+        Assert.That(
+            () => { dispatch = runtime.SendAsync(new FakeCqrsContext(), new ContextAwareRequest()); },
+            Throws.Nothing);
+        Assert.That(
+            async () => await dispatch.ConfigureAwait(false),
+            Throws.InvalidOperationException.With.Message.Contains("does not implement IArchitectureContext"));
+    }
+
+    /// <summary>
+    ///     验证 request handler 缺失时，dispatcher 仍返回 faulted <see cref="ValueTask{TResult}" />，
+    ///     而不是在调用点同步抛出异常。
+    /// </summary>
+    [Test]
+    public void SendAsync_Should_Return_Faulted_ValueTask_When_Handler_Is_Missing()
+    {
+        var runtime = CreateRuntime(
+            container =>
+            {
+                container
+                    .Setup(currentContainer => currentContainer.Get(typeof(IRequestHandler<ContextAwareRequest, int>)))
+                    .Returns((object?)null);
+                container
+                    .Setup(currentContainer => currentContainer.HasRegistration(typeof(IPipelineBehavior<ContextAwareRequest, int>)))
+                    .Returns(false);
+                container
+                    .Setup(currentContainer => currentContainer.GetAll(typeof(IPipelineBehavior<ContextAwareRequest, int>)))
+                    .Returns(Array.Empty<object>());
+            });
+
+        ValueTask<int> dispatch = default;
+        Assert.That(
+            () => { dispatch = runtime.SendAsync(new FakeCqrsContext(), new ContextAwareRequest()); },
+            Throws.Nothing);
+        Assert.That(
+            async () => await dispatch.ConfigureAwait(false),
+            Throws.InvalidOperationException.With.Message.Contains("No CQRS request handler registered"));
+    }
+
     /// <summary>
     ///     验证当 notification handler 需要上下文注入、但当前 CQRS 上下文不实现 <see cref="GFramework.Core.Abstractions.Architectures.IArchitectureContext" /> 时，
     ///     dispatcher 会在发布前显式失败。

GFramework.Cqrs.Tests/Cqrs/CqrsGeneratedRequestInvokerProviderTests.cs

@@ -7,6 +7,7 @@ using GFramework.Core.Architectures;
 using GFramework.Core.Ioc;
 using GFramework.Core.Logging;
 using GFramework.Cqrs.Abstractions.Cqrs;
+using GFramework.Cqrs.Internal;
 
 namespace GFramework.Cqrs.Tests.Cqrs;
 
@@ -99,6 +100,32 @@ internal sealed class CqrsGeneratedRequestInvokerProviderTests
             Is.EqualTo([typeof(GeneratedStreamInvokerProviderRegistry)]));
     }
 
+    /// <summary>
+    ///     验证 direct generated-registry 激活入口只会接入指定 registry，而不会顺手把同一测试程序集里的其他 registry 一并注册。
+    /// </summary>
+    [Test]
+    public void RegisterGeneratedRegistry_Should_Register_Only_The_Selected_Provider()
+    {
+        var container = new MicrosoftDiContainer();
+        var logger = LoggerFactoryResolver.Provider.CreateLogger(nameof(CqrsGeneratedRequestInvokerProviderTests));
+
+        CqrsHandlerRegistrar.RegisterGeneratedRegistry(
+            container,
+            typeof(GeneratedRequestInvokerProviderRegistry),
+            logger);
+
+        var requestProviders = container.GetAll<ICqrsRequestInvokerProvider>();
+        var streamProviders = container.GetAll<ICqrsStreamInvokerProvider>();
+
+        Assert.Multiple(() =>
+        {
+            Assert.That(
+                requestProviders.Select(static provider => provider.GetType()),
+                Is.EqualTo([typeof(GeneratedRequestInvokerProviderRegistry)]));
+            Assert.That(streamProviders, Is.Empty);
+        });
+    }
+
     /// <summary>
     ///     验证当实现类型隐藏、但 stream handler interface 仍可直接表达时，
     ///     registrar 仍会把 generated stream invoker provider 注册到容器中。

GFramework.Cqrs/Internal/CqrsDispatcher.cs

@@ -105,36 +105,43 @@ internal sealed class CqrsDispatcher(
     /// <param name="request">请求对象。</param>
     /// <param name="cancellationToken">取消令牌。</param>
     /// <returns>请求响应。</returns>
-    public async ValueTask<TResponse> SendAsync<TResponse>(
+    public ValueTask<TResponse> SendAsync<TResponse>(
         ICqrsContext context,
         IRequest<TResponse> request,
         CancellationToken cancellationToken = default)
     {
-        ArgumentNullException.ThrowIfNull(context);
-        ArgumentNullException.ThrowIfNull(request);
-
-        var requestType = request.GetType();
-        var dispatchBinding = GetRequestDispatchBinding<TResponse>(requestType);
-        var handler = container.Get(dispatchBinding.HandlerType)
-                      ?? throw new InvalidOperationException(
-                          $"No CQRS request handler registered for {requestType.FullName}.");
-
-        PrepareHandler(handler, context);
-        if (!container.HasRegistration(dispatchBinding.BehaviorType))
+        try
         {
-            return await dispatchBinding.RequestInvoker(handler, request, cancellationToken).ConfigureAwait(false);
+            ArgumentNullException.ThrowIfNull(context);
+            ArgumentNullException.ThrowIfNull(request);
+
+            var requestType = request.GetType();
+            var dispatchBinding = GetRequestDispatchBinding<TResponse>(requestType);
+            var handler = container.Get(dispatchBinding.HandlerType)
+                          ?? throw new InvalidOperationException(
+                              $"No CQRS request handler registered for {requestType.FullName}.");
+
+            PrepareHandler(handler, context);
+            if (!container.HasRegistration(dispatchBinding.BehaviorType))
+            {
+                return dispatchBinding.RequestInvoker(handler, request, cancellationToken);
+            }
+
+            var behaviors = container.GetAll(dispatchBinding.BehaviorType);
+
+            foreach (var behavior in behaviors)
+            {
+                PrepareHandler(behavior, context);
+            }
+
+            return dispatchBinding.GetPipelineExecutor(behaviors.Count)
+                .Invoke(handler, behaviors, request, cancellationToken);
         }
-
-        var behaviors = container.GetAll(dispatchBinding.BehaviorType);
-
-        foreach (var behavior in behaviors)
+        catch (Exception exception)
         {
-            PrepareHandler(behavior, context);
+            // 保留旧 async 实现的 faulted-ValueTask 失败语义，同时继续复用 direct-return 的热路径。
+            return ValueTask.FromException<TResponse>(exception);
         }
-
-        return await dispatchBinding.GetPipelineExecutor(behaviors.Count)
-            .Invoke(handler, behaviors, request, cancellationToken)
-            .ConfigureAwait(false);
     }
 
     /// <summary>

GFramework.Cqrs/Internal/CqrsHandlerRegistrar.cs

@@ -68,6 +68,36 @@ internal static class CqrsHandlerRegistrar
         }
     }
 
+    /// <summary>
+    ///     直接激活并注册单个 generated registry，避免调用方为了只接入一个 benchmark registry
+    ///     而额外扫描同一程序集里的其他 registry / handler。
+    /// </summary>
+    /// <param name="container">承载 generated registry 注册结果的目标容器。</param>
+    /// <param name="registryType">要直接激活的 generated registry 类型。</param>
+    /// <param name="logger">当前注册过程使用的日志记录器。</param>
+    /// <exception cref="ArgumentNullException">
+    ///     <paramref name="container" />、<paramref name="registryType" /> 或 <paramref name="logger" /> 为 <see langword="null" />。
+    /// </exception>
+    /// <exception cref="InvalidOperationException">指定 registry 类型不满足 generated registry 运行时契约。</exception>
+    internal static void RegisterGeneratedRegistry(
+        IIocContainer container,
+        Type registryType,
+        ILogger logger)
+    {
+        ArgumentNullException.ThrowIfNull(container);
+        ArgumentNullException.ThrowIfNull(registryType);
+        ArgumentNullException.ThrowIfNull(logger);
+
+        var assemblyName = GetAssemblySortKey(registryType.Assembly);
+        if (!TryCreateGeneratedRegistry(registryType, assemblyName, logger, out var registry))
+        {
+            throw new InvalidOperationException(
+                $"Unable to activate generated CQRS handler registry {registryType.FullName} in assembly {assemblyName}.");
+        }
+
+        RegisterGeneratedRegistries(container.GetServicesUnsafe, [registry], assemblyName, logger);
+    }
+
     /// <summary>
     ///     优先使用程序集级源码生成注册器完成 CQRS 映射注册。
     /// </summary>

GFramework.Cqrs/Properties/AssemblyInfo.cs

@@ -0,0 +1,7 @@
+// Copyright (c) 2025-2026 GeWuYou
+// SPDX-License-Identifier: Apache-2.0
+
+using System.Runtime.CompilerServices;
+
+[assembly: InternalsVisibleTo("GFramework.Cqrs.Tests")]
+[assembly: InternalsVisibleTo("GFramework.Cqrs.Benchmarks")]

ai-plan/public/cqrs-rewrite/todos/cqrs-rewrite-migration-tracking.md

@@ -7,10 +7,12 @@ CQRS 迁移与收敛。
 
 ## 当前恢复点
 
-- 恢复点编号：`CQRS-REWRITE-RP-103`
+- 恢复点编号：`CQRS-REWRITE-RP-110`
 - 当前阶段：`Phase 8`
-- 当前 PR 锚点：`PR #340`
+- 当前 PR 锚点：`PR #341`
 - 当前结论：
+  - 当前 `RP-110` 已再次使用 `$gframework-pr-review` 复核 `PR #341` latest-head review：`BenchmarkHostFactory` 的 legacy runtime alias 防守式类型检查、benchmark 宿主定向 generated registry 激活、以及 `CqrsDispatcher.SendAsync(...)` 的 faulted `ValueTask` 失败语义在当前 head 均已实质收口；本轮仅继续接受仍然成立的 CodeRabbit nitpick，为 `SendAsync_Should_Return_Faulted_ValueTask_When_Handler_Is_Missing()` 补齐 `HasRegistration(...)` / `GetAll(...)` 防御性 mock，并删除 trace 中重复 `本轮权威验证` 的 `本轮下一步` 段落
+  - 当前 `RP-109` 已使用 `$gframework-pr-review` 复核 `PR #341` latest-head review：benchmark 宿主改为定向激活当前场景的 generated registry，避免同一 benchmark 程序集里的其他 registry 扩大冻结服务索引与 `HasRegistration` 基线；`BenchmarkHostFactory` 为 legacy runtime alias 注册补齐防守式类型检查与 stream lifetime 运行时注释；`CqrsDispatcher.SendAsync(...)` 在保留 direct-return 热路径的同时恢复 faulted `ValueTask` 失败语义，并补齐 generated registry 定向接线与 request fault 语义回归测试；`.agents/skills/gframework-batch-boot/SKILL.md` 的 MD005 缩进也已顺手修正
   - `GFramework.Cqrs` 已完成对外部 `Mediator` 的生产级替代，当前主线已从“是否可替代”转向“仓库内部收口与能力深化顺序”
   - `dispatch/invoker` 生成前移已扩展到 request / stream 路径，`RP-077` 已补齐 request invoker provider gate 与 stream gate 对称的 descriptor / descriptor entry runtime 合同回归
   - `RP-078` 已补齐 mixed fallback metadata 在 runtime 不允许多个 fallback attribute 实例时的单字符串 attribute 回退回归
@@ -40,23 +42,38 @@ CQRS 迁移与收敛。
   - 当前 `RP-102` 已把 `GFramework.Cqrs.Benchmarks` 的 `Mediator` 对照组收口为官方 NuGet 引用（`Mediator.Abstractions` / `Mediator.SourceGenerator` `3.0.2`），不再使用本地 `ai-libs/Mediator` project reference；`RequestBenchmarks` 现已新增 source-generated concrete `Mediator` 对照方法，并通过 `RequestLifetimeBenchmarks` 复核 hot path 收口后的新基线
   - 当前 `RP-102` 已将 `BenchmarkDotNet.Artifacts/` 收口为默认忽略路径，并把 request steady-state / lifetime benchmark 复跑升级为 CQRS 性能相关改动的默认回归门槛；当前阶段目标明确为“持续逼近 source-generated `Mediator`，并至少稳定超过反射版 `MediatR`”
   - 当前 `RP-103` 已使用 `$gframework-pr-review` 复核 `PR #340` latest-head review：修复 `CreateStream_Should_Throw_When_Stream_Pipeline_Behavior_Context_Does_Not_Implement_IArchitectureContext` 因 strict mock 未配置 `HasRegistration(Type)` 产生的 CI 失败，收紧 `MicrosoftDiContainer.HasRegistration(Type)` 到与 `GetAll(Type)` 一致的服务键可见性语义，补齐 `IIocContainer.HasRegistration(Type)` 的异常/XML 契约与 `docs/zh-CN/core/ioc.md` 的用户接入说明，并同步 benchmark 注释与 active tracking/trace 到当前 PR 锚点
-- `ai-plan` active 入口现以 `RP-103` 为最新恢复锚点；`PR #340`、`PR #339`、`PR #334`、`PR #331`、`PR #326`、`PR #323`、`PR #307` 与其他更早阶段细节均以下方归档或说明为准
+  - 当前 `RP-104` 已继续沿用 `$gframework-batch-boot 50` 压 request 热路径：先把 `CqrsDispatcher.SendAsync(...)` 改成 direct-return `ValueTask`，移除 dispatcher 自身的 `async/await` 状态机；再让 `MicrosoftDiContainer.HasRegistration(Type)` 在冻结后复用预构建的服务键索引，避免每次命中零 pipeline request 都线性扫描全部描述符；本轮 benchmark 表明第一刀显著压低 steady-state / lifetime request，第二刀在当前短跑下主要确认“无回退、收益不明显”
+  - 当前 `RP-105` 已继续沿用 `$gframework-batch-boot 50` 压默认 request steady-state：为 benchmark 最小宿主补齐 CQRS runtime / registrar / registration service 基础设施，让 `RequestBenchmarks` 不再只测反射路径，而是通过 handwritten generated registry + `RegisterCqrsHandlersFromAssembly(...)` 真实接上 generated request invoker provider；本轮 benchmark 表明默认 request 路径进一步从约 `70.298 ns / 32 B` 压到约 `65.296 ns / 32 B`，`Singleton / Transient` lifetime 也同步收敛到约 `68.772 ns / 32 B` 与 `73.157 ns / 56 B`
+  - 当前 `RP-106` 已把同一套 generated-provider 宿主收口扩展到 `RequestPipelineBenchmarks`：新增 handwritten `GeneratedRequestPipelineBenchmarkRegistry`，并让 `RequestPipelineBenchmarks` 改走 `RegisterCqrsHandlersFromAssembly(...)` + benchmark CQRS 基础设施预接线；本轮 benchmark 表明 `0 pipeline` steady-state 进一步收敛到约 `64.755 ns / 32 B`，`1 pipeline` 约 `353.141 ns / 536 B`，`4 pipeline` 在短跑噪音下维持约 `555.083 ns / 896 B`
+  - 当前 `RP-107` 已把默认 stream steady-state 宿主也切到 generated-provider 路径：新增 handwritten `GeneratedDefaultStreamingBenchmarkRegistry`，让 `StreamingBenchmarks` 改走 `RegisterCqrsHandlersFromAssembly(...)` 并在 setup/cleanup 清理 dispatcher cache；同时将 `gframework-boot` / `gframework-batch-boot` 的默认停止规则改为“AI 上下文预算优先，建议在预计接近约 80% 安全上下文占用前收口”，不再把 changed files 误当作唯一阈值
+  - 当前 `RP-108` 已补齐 stream handler `Singleton / Transient` 生命周期矩阵 benchmark：新增 `StreamLifetimeBenchmarks` 与 `GeneratedStreamLifetimeBenchmarkRegistry`，让 stream 生命周期对照沿用 generated-provider 宿主接线而不是退回纯反射路径；本轮 benchmark 表明 `Singleton` 下 baseline / `GFramework.Cqrs` / `MediatR` 约 `80.144 ns / 137.515 ns / 229.242 ns`，`Transient` 下约 `77.198 ns / 144.998 ns / 228.185 ns`
+- `ai-plan` active 入口现以 `RP-108` 为最新恢复锚点；`PR #340`、`PR #339`、`PR #334`、`PR #331`、`PR #326`、`PR #323`、`PR #307` 与其他更早阶段细节均以下方归档或说明为准
 
 ## 当前活跃事实
 
 - 当前分支为 `feat/cqrs-optimization`
-- 本轮 `$gframework-batch-boot 50` 以 `origin/main` (`5dc2dd25`, 2026-05-08 09:08:37 +0800) 为基线；本地 `main` (`c2d22285`) 已落后，不作为 branch diff 基线
-- 当前分支相对 `origin/main` 的累计 branch diff 仍为 `10 files / 298 lines`；本轮待提交工作树以 `.gitignore`、benchmark README 与 active tracking/trace 更新为主，仍明显低于 `$gframework-batch-boot 50` 的文件阈值
+- 本轮 `$gframework-batch-boot 50` 以 `origin/main` (`4d6dbba6`, 2026-05-08 11:13:33 +0800) 为基线；本地 `main` 仍落后，不作为 branch diff 基线
+- 当前已提交分支相对 `origin/main` 的累计 branch diff 为 `14 files / 507 lines`
+- 本批待提交工作树集中在 `GFramework.Cqrs.Benchmarks/Messaging/StreamLifetimeBenchmarks.cs`、`GFramework.Cqrs.Benchmarks/Messaging/GeneratedStreamLifetimeBenchmarkRegistry.cs` 与 `GFramework.Cqrs.Benchmarks/README.md`
+- 当前批次后的默认停止依据已改为 AI 上下文预算：若下一轮预计会让活动对话、已加载 recovery 文档、验证输出与当前 diff 接近约 `80%` 安全上下文占用，应在当前自然批次边界停止，即使 branch diff 仍有余量
 - `GFramework.Cqrs.Benchmarks` 作为 benchmark 基础设施项目，必须持续排除在 NuGet / GitHub Packages 发布集合之外
 - `GFramework.Cqrs.Benchmarks` 现已覆盖 request steady-state、pipeline 数量矩阵、startup、request/stream generated invoker，以及 request handler `Singleton / Transient` 生命周期矩阵
 - `GFramework.Cqrs.Benchmarks` 当前以 NuGet 方式引用 `Mediator.Abstractions` / `Mediator.SourceGenerator` `3.0.2`；`ai-libs/Mediator` 只保留为本地源码/README 对照资料，不再参与 benchmark 项目编译
-- 当前 request steady-state benchmark 已形成 baseline / `Mediator` / `MediatR` / `GFramework.Cqrs` 四方对照：约 `5.300 ns / 32 B`、`4.964 ns / 32 B`、`57.993 ns / 232 B`、`83.823 ns / 32 B`
-- 当前 request lifetime benchmark 已从旧坏值显著收敛：`Singleton` 下 `GFramework.Cqrs` 约 `83.183 ns / 32 B`（旧值 `301.731 ns / 440 B`），`Transient` 下约 `86.243 ns / 56 B`（旧值 `287.863 ns / 464 B`）
+- 当前 request steady-state benchmark 已形成 baseline / `Mediator` / `MediatR` / `GFramework.Cqrs` 四方对照：最新约 `5.608 ns / 32 B`、`5.445 ns / 32 B`、`57.071 ns / 232 B`、`64.825 ns / 32 B`
+- 当前 request lifetime benchmark 已继续收敛：`Singleton` 下 `GFramework.Cqrs` 最新约 `69.275 ns / 32 B`，`Transient` 下约 `74.301 ns / 56 B`；相较 `RP-104` 前的 `73.005 ns / 32 B` 与 `74.757 ns / 56 B` 仍维持同一收敛区间
+- 当前 request pipeline benchmark 已改为与默认 request steady-state 相同的 generated-provider 宿主接线路径：`0 pipeline` 约 `64.755 ns / 32 B`，`1 pipeline` 约 `353.141 ns / 536 B`，`4 pipeline` 约 `555.083 ns / 896 B`
+- 当前 stream steady-state benchmark 也已切到 generated-provider 宿主接线路径：baseline 约 `5.535 ns / 32 B`、`MediatR` 约 `59.499 ns / 232 B`、`GFramework.Cqrs` 约 `66.778 ns / 32 B`
+- 当前 stream lifetime benchmark 已补齐 `Singleton / Transient` 两档矩阵，并沿用 generated-provider 宿主接线：`Singleton` 下 baseline / `GFramework.Cqrs` / `MediatR` 约 `80.144 ns / 137.515 ns / 229.242 ns`，`Transient` 下约 `77.198 ns / 144.998 ns / 228.185 ns`
 - 本轮已验证旧 benchmark 劣化的两个主热点：`0 pipeline` 场景下仍解析空行为列表，以及容器查询热路径在 debug 禁用时仍构造日志字符串；两者收口后，`GFramework.Cqrs` request 路径不再出现额外数百字节分配
 - `HasRegistration(Type)` 现在只把“同一服务键已注册”或“开放泛型服务键可闭合到目标类型”视为命中，不再把“仅以具体实现类型自注册”的行为误判为接口服务已注册；该语义与 `Get(Type)` / `GetAll(Type)` 已重新对齐
 - `GFramework.Cqrs.Tests/Cqrs/CqrsDispatcherContextValidationTests.cs` 已同步适配 `HasRegistration(Type)` fast-path，避免 strict mock 因缺少新调用配置而在上下文失败语义断言前提前抛出 `Moq.MockException`
+- `GFramework.Cqrs.Tests/Cqrs/CqrsDispatcherContextValidationTests.cs` 现连“handler 缺失但仍返回 faulted `ValueTask`”这条 request 失败语义回归也显式为 `HasRegistration(Type)` / `GetAll(Type)` 预留了防御性 mock，不再依赖 dispatcher 先判空 handler、后探测 pipeline 的内部顺序
 - `docs/zh-CN/core/ioc.md` 已新增 `HasRegistration(Type)` 的使用语义、热路径用途与“按服务键而非可赋值关系判断”的示例说明
 - 当前 request steady-state 仍落后于 source-generated `Mediator` 与 `MediatR`，但差距已从“额外数百字节分配 + 近 300ns”收敛到“零 pipeline fast-path 仍慢约 `31ns` / `3.6x` 于 `Mediator`”；下一批若继续压 request dispatch，应优先评估默认路径吸收 generated invoker/provider 的空间
+- 本轮 `SendAsync(...)` 的 direct-return `ValueTask` 改动已证明确实是有效热点：同样的短跑配置下，`GFramework.Cqrs` steady-state request 从约 `83.823 ns` 下探到 `69-70 ns` 区间
+- 冻结后 `HasRegistration(Type)` 服务键索引化在当前短跑下没有带来同等量级的可见收益，但也没有引入功能回退或额外分配；后续若继续压零 pipeline request，应优先重新评估“默认 request 路径进一步吸收 generated invoker/provider”而不是继续堆叠同层级微优化
+- 默认 `RequestBenchmarks`、`RequestPipelineBenchmarks` 与 `StreamingBenchmarks` 现在都已通过 handwritten generated registry + 真实 `RegisterCqrsHandlersFromAssembly(...)` 宿主接线命中 generated invoker provider，不再只代表纯反射 binding 路径
+- `gframework-boot` 与 `gframework-batch-boot` 现明确把“上下文预算接近约 80%”视为默认优先停止信号，branch diff files / lines 仅保留为次级仓库范围指标
 - 当前性能回归门槛已收紧为：只要改动触达 `GFramework.Cqrs` request dispatch、DI 热路径、invoker/provider、pipeline 或 benchmark 宿主，就必须至少复跑 `RequestBenchmarks.SendRequest_*` 与 `RequestLifetimeBenchmarks.SendRequest_*`
 - 当前阶段的性能验收目标已明确为：默认 request steady-state 路径不要求超过 source-generated `Mediator`，但必须持续逼近它，并至少稳定快于基于反射 / 扫描的 `MediatR`
 - `GFramework.Core` 当前已通过内部 bridge request / handler 把 legacy `ICommand`、`IAsyncCommand`、`IQuery`、`IAsyncQuery` 接到统一 `ICqrsRuntime`
@@ -98,6 +115,7 @@ CQRS 迁移与收敛。
 - `RequestStartupBenchmarks` 为了量化真正的单次 cold-start，引入了 `InvocationCount=1` / `UnrollFactor=1` 的专用 job；该配置会触发 BenchmarkDotNet 的 `MinIterationTime` 提示，后续若要做稳定基线比较，还需要决定是否引入批量外层循环或自定义 cold-start harness
 - 当前 benchmark 宿主仍刻意保持“单根容器最小宿主”模型；若要公平比较 `Scoped` handler 生命周期，需要先引入显式 scope 创建与 scope 内首次解析的对照基线
 - 当前 `Mediator` 对照组仅先接入 steady-state request；若要把 `Transient` / `Scoped` 生命周期矩阵也纳入同一组对照，需要按 `Mediator` 官方 benchmark 的做法拆分 compile-time lifetime build config，而不是在同一编译产物里混用多个 lifetime
+- 当前 stream 生命周期矩阵尚未接入 `Mediator` concrete runtime；若要继续对齐 `Mediator` 官方 benchmark 的 compile-time lifetime 设计，需要为 stream 场景补专门的 build-time 配置，而不是在当前统一宿主里临时拼接
 - `BenchmarkDotNet.Artifacts/` 现已加入仓库忽略规则；若后续确实需要提交新的基准报告，应显式挑选结果文件或改走文档归档，而不是直接纳入整个生成目录
 - 当前 `GFramework.Cqrs` request steady-state 仍慢于 `MediatR`；在“至少超过反射版 `MediatR`”这个阶段目标达成前，任何相关改动都不能只看功能 build/test 结果，必须附带 benchmark 回归数据
 - 仓库内部仍保留旧 `Command` / `Query` API、`LegacyICqrsRuntime` alias 与部分历史命名语义，后续若不继续分批收口，容易混淆“对外替代已完成”与“内部收口未完成”
@@ -110,6 +128,19 @@ CQRS 迁移与收敛。
 
 ## 最近权威验证
 
+- `dotnet build GFramework.Cqrs.Tests/GFramework.Cqrs.Tests.csproj -c Release`
+  - 结果：通过，`0 warning / 0 error`
+  - 备注：并行验证首轮曾因 `build` 与 `test` 同时写入同一输出 DLL 触发 `MSB3026` 单次复制重试；改为串行重跑同一命令后稳定通过
+- `dotnet test GFramework.Cqrs.Tests/GFramework.Cqrs.Tests.csproj -c Release --filter "FullyQualifiedName~CqrsDispatcherContextValidationTests"`
+  - 结果：通过，`6/6` passed
+- `python3 scripts/license-header.py --check --paths GFramework.Cqrs.Tests/Cqrs/CqrsDispatcherContextValidationTests.cs`
+  - 结果：通过
+- `git diff --check`
+  - 结果：通过
+  - 备注：仅剩 `GFramework.sln` 的历史 CRLF 提示，无本轮新增 diff 格式问题
+- `python3 .agents/skills/gframework-pr-review/scripts/fetch_current_pr_review.py --format json --json-output /tmp/current-pr-review.json`
+  - 结果：通过
+  - 备注：确认当前分支对应 `PR #341`；latest-head 当前仍显示 `CodeRabbit 2` / `Greptile 2` open thread，但其中 `BenchmarkHostFactory` / benchmark registry / faulted `ValueTask` 三类运行时 thread 已在本地失效，当前仅剩测试 mock 脆弱性与 trace 冗余仍值得继续收口
 - `python3 .agents/skills/gframework-pr-review/scripts/fetch_current_pr_review.py --json-output /tmp/current-pr-review.json`
   - 结果：通过
   - 备注：确认当前分支对应 `PR #340`；latest-head 当前显示 `CodeRabbit 2` / `Greptile 2` open thread，且 `CTRF` 报告中唯一失败测试为 `CreateStream_Should_Throw_When_Stream_Pipeline_Behavior_Context_Does_Not_Implement_IArchitectureContext`
@@ -135,9 +166,55 @@ CQRS 迁移与收敛。
   - 备注：按新性能回归门槛复跑后，`Singleton` 下 `GFramework.Cqrs` / `MediatR` 约 `83.183 ns / 32 B` vs `60.915 ns / 232 B`；`Transient` 下约 `86.243 ns / 56 B` vs `59.644 ns / 232 B`
 - `env GIT_DIR=... GIT_WORK_TREE=... python3 scripts/license-header.py --check`
   - 结果：通过
+- `dotnet build GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.csproj -c Release`
+  - 结果：通过，`0 warning / 0 error`
+- `python3 scripts/license-header.py --check --paths GFramework.Cqrs.Benchmarks/Messaging/GeneratedStreamLifetimeBenchmarkRegistry.cs GFramework.Cqrs.Benchmarks/Messaging/StreamLifetimeBenchmarks.cs GFramework.Cqrs.Benchmarks/README.md`
+  - 结果：通过
+- `git diff --check`
+  - 结果：通过
+- `dotnet run --project GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.csproj -c Release -- --filter "*RequestBenchmarks.SendRequest_*" --job short --warmupCount 1 --iterationCount 1 --launchCount 1`
+  - 结果：通过
+  - 备注：steady-state request 对照约为 baseline `5.336 ns / 32 B`、`Mediator` `5.564 ns / 32 B`、`MediatR` `53.307 ns / 232 B`、`GFramework.Cqrs` `64.745 ns / 32 B`
+- `dotnet run --project GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.csproj -c Release -- --filter "*RequestLifetimeBenchmarks.SendRequest_*" --job short --warmupCount 1 --iterationCount 1 --launchCount 1`
+  - 结果：通过
+  - 备注：`Singleton` 下 baseline / `MediatR` / `GFramework.Cqrs` 约 `4.309 ns / 51.923 ns / 67.981 ns`；`Transient` 下约 `5.029 ns / 54.435 ns / 76.437 ns`
+- `dotnet run --project GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.csproj -c Release -- --filter "*StreamLifetimeBenchmarks*" --job short --warmupCount 1 --iterationCount 1 --launchCount 1`
+  - 结果：通过
+  - 备注：`Singleton` 下 baseline / `GFramework.Cqrs` / `MediatR` 约 `80.144 ns / 137.515 ns / 229.242 ns`，`Transient` 下约 `77.198 ns / 144.998 ns / 228.185 ns`
 - `git diff --check`
   - 结果：通过
   - 备注：当前仅保留 `GFramework.sln` 的历史 CRLF 警告，无本轮新增 diff 格式错误
+- `dotnet build GFramework.Cqrs/GFramework.Cqrs.csproj -c Release`
+  - 结果：通过，`0 warning / 0 error`
+- `dotnet test GFramework.Core.Tests/GFramework.Core.Tests.csproj -c Release --filter "FullyQualifiedName~MicrosoftDiContainerTests"`
+  - 结果：通过，`52/52` passed
+- `dotnet test GFramework.Cqrs.Tests/GFramework.Cqrs.Tests.csproj -c Release --no-build --filter "FullyQualifiedName~CqrsDispatcherCacheTests|FullyQualifiedName~CqrsDispatcherContextValidationTests"`
+  - 结果：通过，`14/14` passed
+- `dotnet build GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.csproj -c Release`
+  - 结果：通过，`0 warning / 0 error`
+- `dotnet run --project GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.csproj -c Release --no-build -- --filter "*RequestBenchmarks.SendRequest_*" --job short --warmupCount 1 --iterationCount 1 --launchCount 1`
+  - 结果：通过
+  - 备注：本轮两批热路径收口后的最新 steady-state request 对照约为 baseline `6.141 ns / 32 B`、`Mediator` `6.674 ns / 32 B`、`MediatR` `61.803 ns / 232 B`、`GFramework.Cqrs` `70.298 ns / 32 B`
+- `dotnet run --project GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.csproj -c Release --no-build -- --filter "*RequestLifetimeBenchmarks.SendRequest_*" --job short --warmupCount 1 --iterationCount 1 --launchCount 1`
+  - 结果：通过
+  - 备注：最新 lifetime request 对照约为 `Singleton` 下 baseline / `MediatR` / `GFramework.Cqrs` = `4.706 ns / 52.197 ns / 73.005 ns`，`Transient` 下 = `4.571 ns / 50.175 ns / 74.757 ns`
+- `dotnet build GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.csproj -c Release`
+  - 结果：通过，`0 warning / 0 error`
+- `python3 scripts/license-header.py --check --paths GFramework.Cqrs.Benchmarks/Messaging/GeneratedDefaultRequestBenchmarkRegistry.cs GFramework.Cqrs.Benchmarks/Messaging/BenchmarkHostFactory.cs GFramework.Cqrs.Benchmarks/Messaging/RequestBenchmarks.cs GFramework.Cqrs.Benchmarks/README.md`
+  - 结果：通过
+- `git diff --check`
+  - 结果：通过
+- `dotnet run --project GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.csproj -c Release -- --filter "*RequestBenchmarks.SendRequest_*" --job short --warmupCount 1 --iterationCount 1 --launchCount 1`
+  - 结果：通过
+  - 备注：默认 steady-state request 对照现约为 baseline `5.013 ns / 32 B`、`Mediator` `5.747 ns / 32 B`、`MediatR` `51.588 ns / 232 B`、`GFramework.Cqrs` `65.296 ns / 32 B`
+- `dotnet run --project GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.csproj -c Release -- --filter "*RequestLifetimeBenchmarks.SendRequest_*" --job short --warmupCount 1 --iterationCount 1 --launchCount 1`
+  - 结果：通过
+  - 备注：最新 lifetime request 对照约为 `Singleton` 下 baseline / `MediatR` / `GFramework.Cqrs` = `4.817 ns / 48.177 ns / 68.772 ns`，`Transient` 下 = `4.841 ns / 51.753 ns / 73.157 ns`
+- `env GIT_DIR=... GIT_WORK_TREE=... python3 scripts/license-header.py --check`
+  - 结果：通过
+- `git diff --check`
+  - 结果：通过
+  - 备注：仍仅保留 `GFramework.sln` 的历史 CRLF 警告，无本轮新增 diff 格式问题
 - `dotnet pack GFramework.sln -c Release --no-restore -o /tmp/gframework-pack-validation -p:IncludeSymbols=false`
   - 结果：通过
   - 备注：当前本地产物仅包含 14 个预期发布包，未生成 `GFramework.Cqrs.Benchmarks.*.nupkg`
@@ -258,9 +335,9 @@ CQRS 迁移与收敛。
 
 ## 下一推荐步骤
 
-1. 若继续沿用 `$gframework-batch-boot 50` 且优先处理性能，下一批先对 `CqrsDispatcher.SendAsync(...)` / request invoker 绑定 / handler 调用适配做更细粒度热点拆分，并在每次改动后立即复跑 `RequestBenchmarks` 与 `RequestLifetimeBenchmarks`
-2. 若要把“至少超过反射版 `MediatR`”变成可执行目标，下一批优先评估默认 request 路径吸收 generated invoker/provider 或继续裁掉 dispatch binding / delegate 适配层的剩余常量开销
-3. 若 benchmark 对照需要继续贴近 `Mediator` 官方设计，再扩 `Mediator` 的 compile-time lifetime 矩阵，而不是先横向堆更多低价值场景
+1. 当前 turn 已到新的自然批次边界；本次提交后应停止，并在新的 turn 里从 `RP-108` 恢复点继续，而不是在本轮继续启动新的 benchmark 宿主或 runtime 热点切片
+2. 若下一轮继续沿用 `$gframework-batch-boot` 且优先处理性能，先看 notification publish 或更高价值的 request dispatch 常量开销热点，而不是继续堆同层级 benchmark 宿主补齐
+3. 若 benchmark 对照需要继续贴近 `Mediator` 官方设计，再评估 `Mediator` 的 compile-time lifetime / stream 对照矩阵，或给 stream 引入 scoped host 基线，而不是回头重试已被 benchmark 否决的 `GetAll(Type)` 零行为探测方案
 
 ## 活跃文档
 

ai-plan/public/cqrs-rewrite/traces/cqrs-rewrite-migration-trace.md

@@ -2,6 +2,243 @@
 
 ## 2026-05-08
 
+### 阶段：PR #341 latest-head review 尾声收口（CQRS-REWRITE-RP-110）
+
+- 再次使用 `$gframework-pr-review` 抓取 `PR #341` latest-head review，确认当前 open thread 已收敛到：
+  - `BenchmarkHostFactory.cs` 的 legacy runtime alias 防守式类型检查 thread，但当前 head 已存在 `RegisterLegacyRuntimeAlias(...)` 的显式类型校验与实际类型信息异常，属于 GitHub 未 resolve 的 stale thread
+  - `RequestBenchmarks.cs` / `CqrsDispatcher.cs` 的 Greptile thread，对应“程序集级 registry 扩散”与“faulted ValueTask 失败语义”均已在当前 head 修复，属于 stale thread
+  - 仍然成立且值得当前收口的只剩 `CqrsDispatcherContextValidationTests.cs` 的 strict mock 脆弱性，以及本 trace 中 `本轮下一步` 与 `本轮权威验证` 重复的问题
+- 本轮主线程决策：
+  - 为 `SendAsync_Should_Return_Faulted_ValueTask_When_Handler_Is_Missing()` 补齐 `HasRegistration(...)` 与 `GetAll(...)` 的防御性 mock，降低该测试对 dispatcher 内部检查顺序的隐式耦合
+  - 删除 `RP-109` 记录中重复 `本轮权威验证` 的 `本轮下一步` 段落，保持默认恢复入口只保留仍有价值的恢复信息
+- 本轮权威验证：
+  - `python3 .agents/skills/gframework-pr-review/scripts/fetch_current_pr_review.py --format json --json-output /tmp/current-pr-review.json`
+    - 结果：通过
+    - 备注：确认当前分支对应 `PR #341`；latest-head 当前仍显示 `CodeRabbit 2` / `Greptile 2` open thread，但其中运行时/benchmark 两条已在本地失效
+  - `dotnet build GFramework.Cqrs.Tests/GFramework.Cqrs.Tests.csproj -c Release`
+    - 结果：通过，`0 warning / 0 error`
+    - 备注：并行验证首轮曾因 `build` 与 `test` 同时写入同一输出 DLL 触发 `MSB3026` 单次复制重试；改为串行重跑同一命令后稳定通过
+  - `dotnet test GFramework.Cqrs.Tests/GFramework.Cqrs.Tests.csproj -c Release --filter "FullyQualifiedName~CqrsDispatcherContextValidationTests"`
+    - 结果：通过，`6/6` passed
+  - `python3 scripts/license-header.py --check --paths GFramework.Cqrs.Tests/Cqrs/CqrsDispatcherContextValidationTests.cs`
+    - 结果：通过
+  - `git diff --check`
+    - 结果：通过
+    - 备注：仅剩 `GFramework.sln` 的历史 CRLF 提示，无本轮新增 diff 格式问题
+
+### 阶段：PR #341 latest-head review 收口（CQRS-REWRITE-RP-109）
+
+- 使用 `$gframework-pr-review` 抓取 `feat/cqrs-optimization` 当前公开 PR，并确认当前锚点已从 `PR #340` 更新为 `PR #341`
+- 本轮 latest-head review 结论：
+  - `CodeRabbit` 仍有 `BenchmarkHostFactory.cs` 的 legacy runtime 硬转型、`StreamLifetimeBenchmarks.cs` 的注释缺口，以及 `.agents/skills/gframework-batch-boot/SKILL.md` 的 `MD005` 缩进问题
+  - `Greptile` 指出的两条仍然成立：benchmark 项目里通过 `RegisterCqrsHandlersFromAssembly(typeof(...).Assembly)` 会把同程序集的其他 generated registry 一并激活，扩大 benchmark 宿主的服务索引基线；`CqrsDispatcher.SendAsync(...)` 直接去掉 `async/await` 后也把原本的 faulted-`ValueTask` 失败语义改成了同步抛出
+- 本轮主线程决策：
+  - 在 `GFramework.Cqrs.Internal.CqrsHandlerRegistrar` 新增 direct generated-registry 激活入口，并通过 `InternalsVisibleTo` 暴露给 `GFramework.Cqrs.Benchmarks`，让 benchmark 宿主只激活当前场景的 generated registry
+  - 把 `RequestBenchmarks`、`RequestPipelineBenchmarks`、`StreamingBenchmarks`、`StreamLifetimeBenchmarks` 以及 request/stream invoker benchmark 的 generated 宿主全部切到定向 registry 接线，避免同程序集其他 registry 扩大冻结索引和 descriptor 预热基线
+  - 在 `BenchmarkHostFactory` 里用防守式类型检查注册 legacy runtime alias，并补充 stream lifetime runtime 二次创建的注释
+  - 让 `CqrsDispatcher.SendAsync(...)` 通过 `ValueTask.FromException<TResponse>(...)` 恢复旧的 faulted-`ValueTask` 失败语义，同时保留成功路径的 direct-return 热路径
+  - 补齐 `CqrsGeneratedRequestInvokerProviderTests` 与 `CqrsDispatcherContextValidationTests` 的 targeted 回归，并顺手修正 batch boot skill 的 markdown 缩进
+- 本轮权威验证：
+  - `dotnet build GFramework.Cqrs/GFramework.Cqrs.csproj -c Release`
+    - 结果：通过，`1 warning / 0 error`
+    - 备注：仅出现 `MSB3026` 单次复制重试告警，随后成功产出 `net10.0` 目标；未出现编译失败或新增代码警告
+  - `dotnet build GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.csproj -c Release`
+    - 结果：通过，`0 warning / 0 error`
+  - `dotnet test GFramework.Cqrs.Tests/GFramework.Cqrs.Tests.csproj -c Release --filter "FullyQualifiedName~CqrsDispatcherContextValidationTests|FullyQualifiedName~CqrsGeneratedRequestInvokerProviderTests"`
+    - 结果：通过，`24/24` passed
+    - 备注：首轮并行验证时因与 build 同时运行触发 MSBuild 输出文件锁竞争；改为串行重跑同一命令后稳定通过
+  - `python3 scripts/license-header.py --check --paths GFramework.Cqrs/Properties/AssemblyInfo.cs GFramework.Cqrs/Internal/CqrsHandlerRegistrar.cs GFramework.Cqrs/Internal/CqrsDispatcher.cs GFramework.Cqrs.Benchmarks/Messaging/BenchmarkHostFactory.cs GFramework.Cqrs.Benchmarks/Messaging/RequestBenchmarks.cs GFramework.Cqrs.Benchmarks/Messaging/RequestPipelineBenchmarks.cs GFramework.Cqrs.Benchmarks/Messaging/StreamingBenchmarks.cs GFramework.Cqrs.Benchmarks/Messaging/StreamLifetimeBenchmarks.cs GFramework.Cqrs.Benchmarks/Messaging/RequestInvokerBenchmarks.cs GFramework.Cqrs.Benchmarks/Messaging/StreamInvokerBenchmarks.cs GFramework.Cqrs.Tests/Cqrs/CqrsDispatcherContextValidationTests.cs GFramework.Cqrs.Tests/Cqrs/CqrsGeneratedRequestInvokerProviderTests.cs`
+    - 结果：通过
+    - 备注：仓库脚本默认内部调用未绑定 worktree 的 `git ls-files`，因此本轮按修改文件列表显式 `--paths` 校验
+  - `git diff --check`
+    - 结果：通过
+
+### 阶段：stream handler 生命周期矩阵 benchmark（CQRS-REWRITE-RP-108）
+
+- 延续 `$gframework-batch-boot 50`，本轮继续使用 `origin/main` 作为 branch diff 基线，并先复核：
+  - `origin/main` = `4d6dbba6`，提交时间 `2026-05-08 11:13:33 +0800`
+  - 当前分支 `feat/cqrs-optimization` 相对 `origin/main` 的累计 branch diff 为 `14 files / 507 lines`
+  - 当前 turn 虽然仍低于 `50 files` 阈值，但已加载多轮 recovery / benchmark 输出；因此只允许再推进一个单模块、低风险 benchmark 切片
+- 本轮接受的只读探索结论：
+  - `RequestLifetimeBenchmarks` 已覆盖 request 的 `Singleton / Transient` 生命周期矩阵，但 stream 侧仍缺少对称的 handler 生命周期对照
+  - `StreamingBenchmarks` 已在 `RP-107` 切到 generated-provider 宿主，适合作为 stream 生命周期矩阵的宿主基础；继续退回纯反射路径会让“生命周期变量”和“descriptor 路径变量”混在一起
+  - 如果让 generated registry 顺手注册默认单例 handler，会破坏生命周期矩阵的变量控制，因此 registry 只能暴露 descriptor，不能抢先锁死 handler 生命周期
+- 本轮主线程决策：
+  - 新增 `StreamLifetimeBenchmarks`，对齐 request 生命周期矩阵，只比较 `Singleton / Transient` 两档，继续明确把 `Scoped` 留给未来显式 scoped host
+  - 新增 `GeneratedStreamLifetimeBenchmarkRegistry`，只提供 handwritten generated stream invoker descriptor，不直接注册 handler
+  - 让 `StreamLifetimeBenchmarks` 使用 `RegisterCqrsHandlersFromAssembly(typeof(StreamLifetimeBenchmarks).Assembly)` 建立 generated-provider 宿主，再显式按 benchmark 参数注册 `Singleton / Transient` handler 生命周期
+  - 更新 `GFramework.Cqrs.Benchmarks/README.md`，把 stream 生命周期矩阵列为已覆盖场景，并从“后续扩展方向”里移除这项待办
+- 本轮验证过程的重要补充：
+  - 首次并行触发 `RequestBenchmarks` / `RequestLifetimeBenchmarks` / `StreamLifetimeBenchmarks` 时，在同一 autogenerated BenchmarkDotNet 目录下复现了文件已存在冲突与 bootstrap 异常；这是 benchmark 基础设施层面的并行目录竞争，不是代码缺陷
+  - 改为串行重跑后三组 benchmark 全部稳定通过，因此本轮将“BenchmarkDotNet 在当前仓库里不应并行运行多条 `dotnet run --project ... --filter ...` 会话”视为有效执行约束
+- 本轮权威验证：
+  - `dotnet build GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.csproj -c Release`
+    - 结果：通过，`0 warning / 0 error`
+  - `python3 scripts/license-header.py --check --paths GFramework.Cqrs.Benchmarks/Messaging/GeneratedStreamLifetimeBenchmarkRegistry.cs GFramework.Cqrs.Benchmarks/Messaging/StreamLifetimeBenchmarks.cs GFramework.Cqrs.Benchmarks/README.md`
+    - 结果：通过
+  - `git diff --check`
+    - 结果：通过
+  - `dotnet run --project GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.csproj -c Release -- --filter "*RequestBenchmarks.SendRequest_*" --job short --warmupCount 1 --iterationCount 1 --launchCount 1`
+    - 结果：通过
+    - 备注：steady-state request 对照约为 baseline `5.336 ns / 32 B`、`Mediator` `5.564 ns / 32 B`、`MediatR` `53.307 ns / 232 B`、`GFramework.Cqrs` `64.745 ns / 32 B`
+  - `dotnet run --project GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.csproj -c Release -- --filter "*RequestLifetimeBenchmarks.SendRequest_*" --job short --warmupCount 1 --iterationCount 1 --launchCount 1`
+    - 结果：通过
+    - 备注：`Singleton` 下 baseline / `MediatR` / `GFramework.Cqrs` 约 `4.309 ns / 51.923 ns / 67.981 ns`；`Transient` 下约 `5.029 ns / 54.435 ns / 76.437 ns`
+  - `dotnet run --project GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.csproj -c Release -- --filter "*StreamLifetimeBenchmarks*" --job short --warmupCount 1 --iterationCount 1 --launchCount 1`
+    - 结果：通过
+    - 备注：`Singleton` 下 baseline / `GFramework.Cqrs` / `MediatR` 约 `80.144 ns / 137.515 ns / 229.242 ns`，`Transient` 下约 `77.198 ns / 144.998 ns / 228.185 ns`
+- 本轮结论：
+  - stream 生命周期矩阵现在已与 request 生命周期矩阵对称，且继续沿用 generated-provider 宿主路径，没有把变量退化回纯反射 binding
+  - `GFramework.Cqrs` 在 stream `Singleton / Transient` 两档下都明显快于 `MediatR`，同时保持接近 baseline 的分配规模；`Transient` 仅从 `240 B` 小幅增至 `264 B`
+  - 真正的停止依据仍是上下文预算安全。虽然 branch diff 只有 `14 files`，但当前 turn 已包含多轮 benchmark 输出和恢复文档，因此本批提交后应主动停止
+  - 下一轮若继续性能线，更值得优先看 notification publish 或更高价值的 request 常量开销热点，而不是继续做同层级 benchmark 宿主补齐
+
+### 阶段：默认 stream benchmark 吸收 generated provider 宿主（CQRS-REWRITE-RP-107）
+
+- 延续 `$gframework-batch-boot 50`，但本轮按用户新增要求把默认停止依据改为“AI 上下文预算优先，建议在预计接近约 80% 安全上下文占用前收口”；在真正落代码前先复核：
+  - `origin/main` = `4d6dbba6`，提交时间 `2026-05-08 11:13:33 +0800`
+  - 当前分支 `feat/cqrs-optimization` 相对 `origin/main` 的累计 branch diff 为 `10 files / 507 lines`
+  - 当前 turn 已加载 `AGENTS.md`、`gframework-batch-boot` / `gframework-boot`、active tracking/trace、上一轮 benchmark 结果与多次 validation 输出，因此继续一个自然批次可以接受，但不应在本次提交后继续无界循环
+- 本轮接受的只读探索结论：
+  - 默认 request / request pipeline 宿主都已吸收 generated provider，但 `StreamingBenchmarks` 仍停在“直接注册单个 stream handler”的旧宿主路径，口径与 `StreamInvokerBenchmarks` / 默认 request 组不对称
+  - 默认 stream steady-state 场景已经足够独立，适合用一份新的 handwritten generated stream registry 最小化收口，而不用再修改 runtime 语义
+  - 用户要求把停止条件从 changed files 改成 AI 上下文预算，因此 skill 文档本身也属于这一批必须一起落下的恢复边界更新
+- 本轮主线程决策：
+  - 新增 `GeneratedDefaultStreamingBenchmarkRegistry`，用 handwritten generated registry + `ICqrsStreamInvokerProvider` + `IEnumeratesCqrsStreamInvokerDescriptors` 为 `StreamingBenchmarks.BenchmarkStreamRequest` 提供真实的 generated stream invoker descriptor
+  - 让 `StreamingBenchmarks` 改用 `RegisterCqrsHandlersFromAssembly(typeof(StreamingBenchmarks).Assembly)` 建容器，并在 `Setup/Cleanup` 前后显式清理 dispatcher 静态缓存
+  - 更新 `GFramework.Cqrs.Benchmarks/README.md`，明确默认 stream steady-state benchmark 也已接上 handwritten generated stream invoker provider
+  - 更新 `.agents/skills/gframework-batch-boot/SKILL.md` 与 `.agents/skills/gframework-boot/SKILL.md`，明确“上下文预算接近约 80% 时优先停止，branch diff 文件/行数只作次级仓库范围信号”
+- 本轮权威验证：
+  - `dotnet build GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.csproj -c Release`
+    - 结果：通过，`0 warning / 0 error`
+  - `python3 scripts/license-header.py --check --paths GFramework.Cqrs.Benchmarks/Messaging/GeneratedDefaultStreamingBenchmarkRegistry.cs GFramework.Cqrs.Benchmarks/Messaging/StreamingBenchmarks.cs GFramework.Cqrs.Benchmarks/README.md`
+    - 结果：通过
+  - `git diff --check`
+    - 结果：通过
+  - `dotnet run --project GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.csproj -c Release -- --filter "*RequestBenchmarks.SendRequest_*" --job short --warmupCount 1 --iterationCount 1 --launchCount 1`
+    - 结果：通过
+    - 备注：steady-state request 对照约为 baseline `5.608 ns / 32 B`、`Mediator` `5.445 ns / 32 B`、`MediatR` `57.071 ns / 232 B`、`GFramework.Cqrs` `64.825 ns / 32 B`
+  - `dotnet run --project GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.csproj -c Release -- --filter "*RequestLifetimeBenchmarks.SendRequest_*" --job short --warmupCount 1 --iterationCount 1 --launchCount 1`
+    - 结果：通过
+    - 备注：`Singleton` 下 baseline / `MediatR` / `GFramework.Cqrs` 约 `4.446 ns / 51.331 ns / 69.275 ns`；`Transient` 下约 `4.918 ns / 56.382 ns / 74.301 ns`
+  - `dotnet run --project GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.csproj -c Release -- --filter "*StreamingBenchmarks*" --job short --warmupCount 1 --iterationCount 1 --launchCount 1`
+    - 结果：通过
+    - 备注：默认 stream steady-state 对照约为 baseline `5.535 ns / 32 B`、`MediatR` `59.499 ns / 232 B`、`GFramework.Cqrs` `66.778 ns / 32 B`
+- 本轮结论：
+  - 默认 stream steady-state benchmark 现在也已切到 generated-provider 宿主路径，request / pipeline / stream 三个默认宿主场景的 benchmark 口径终于对齐
+  - `StreamingBenchmarks` 的 `GFramework.Cqrs` 结果约 `66.778 ns / 32 B`，仍慢于 `MediatR`，但没有新增分配或明显回退，说明这次宿主收口是低风险可接受的
+  - 更重要的是，默认停止依据已从“branch diff 文件数是否触顶”改成“AI 上下文预算是否接近约 80%”；结合当前 turn 已加载的大量 recovery/validation/benchmark 输出，本次提交后应主动停止，而不是继续机械扩批
+  - 下一轮若继续性能线，应从 `RP-107` 恢复点重新进入，并优先挑选新的高价值热点族，而不是沿着当前 turn 再追加更多同类宿主收口
+
+### 阶段：request pipeline benchmark 吸收 generated provider 宿主（CQRS-REWRITE-RP-106）
+
+- 延续 `$gframework-batch-boot 50`，本轮基于 `RP-105` 已验证的默认 request 宿主接线继续推进，并先复核 branch diff 基线：
+  - `origin/main` = `4d6dbba6`，提交时间 `2026-05-08 11:13:33 +0800`
+  - 当前分支 `feat/cqrs-optimization` 相对 `origin/main` 的累计 branch diff 为 `8 files / 358 lines`
+  - 当前工作树待提交改动只集中在 `RequestPipelineBenchmarks`、对应 handwritten generated registry 与 benchmark `README`，因此继续自动推进下一批 pipeline 宿主收口
+- 本轮接受的只读探索结论：
+  - `RP-105` 已证明“让默认 request 宿主真实接上 generated request invoker provider”能稳定压低 steady-state request，因此 pipeline benchmark 仍保留旧的“直接注册单个 handler”路径会让口径不对齐
+  - 之前已被 benchmark 否决的“总是 `GetAll(Type)` 做零 pipeline 探测”不应回头重试；下一刀更合理的是把 pipeline benchmark 也切到真实程序集注册入口
+  - `RequestPipelineBenchmarks` 只需要补一份与 `RequestBenchmarks` 对称的 handwritten generated registry，就能最小化改动并保持 runtime 语义不变
+- 本轮主线程决策：
+  - 新增 `GeneratedRequestPipelineBenchmarkRegistry`，用 handwritten generated registry + `ICqrsRequestInvokerProvider` + `IEnumeratesCqrsRequestInvokerDescriptors` 为 `RequestPipelineBenchmarks.BenchmarkRequest` 提供真实的 generated request invoker descriptor
+  - 让 `RequestPipelineBenchmarks` 改用 `RegisterCqrsHandlersFromAssembly(typeof(RequestPipelineBenchmarks).Assembly)` 建容器，只把 pipeline 行为数量矩阵保留在 benchmark 自己的显式注册里
+  - 更新 `GFramework.Cqrs.Benchmarks/README.md`，明确 request pipeline benchmark 也已接上 handwritten generated request invoker provider
+- 本轮权威验证：
+  - `dotnet build GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.csproj -c Release`
+    - 结果：通过，`0 warning / 0 error`
+  - `python3 scripts/license-header.py --check --paths GFramework.Cqrs.Benchmarks/Messaging/GeneratedRequestPipelineBenchmarkRegistry.cs GFramework.Cqrs.Benchmarks/Messaging/RequestPipelineBenchmarks.cs GFramework.Cqrs.Benchmarks/README.md`
+    - 结果：通过
+  - `git diff --check`
+    - 结果：通过
+  - `dotnet run --project GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.csproj -c Release -- --filter "*RequestBenchmarks.SendRequest_*" --job short --warmupCount 1 --iterationCount 1 --launchCount 1`
+    - 结果：通过
+    - 备注：steady-state request 对照约为 baseline `5.680 ns / 32 B`、`Mediator` `6.565 ns / 32 B`、`MediatR` `54.737 ns / 232 B`、`GFramework.Cqrs` `63.644 ns / 32 B`
+  - `dotnet run --project GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.csproj -c Release -- --filter "*RequestLifetimeBenchmarks.SendRequest_*" --job short --warmupCount 1 --iterationCount 1 --launchCount 1`
+    - 结果：通过
+    - 备注：`Singleton` 下 `GFramework.Cqrs` / `MediatR` 约 `69.896 ns / 32 B` vs `57.469 ns / 232 B`；`Transient` 下约 `72.880 ns / 56 B` vs `55.106 ns / 232 B`
+  - `dotnet run --project GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.csproj -c Release -- --filter "*RequestPipelineBenchmarks.SendRequest_GFrameworkCqrs*" --job short --warmupCount 1 --iterationCount 1 --launchCount 1`
+    - 结果：通过
+    - 备注：第一次短跑为 `PipelineCount=0` `64.928 ns / 32 B`、`PipelineCount=1` `366.468 ns / 536 B`、`PipelineCount=4` `547.800 ns / 896 B`
+  - `dotnet run --project GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.csproj -c Release -- --filter "*RequestPipelineBenchmarks.SendRequest_GFrameworkCqrs*" --job short --warmupCount 1 --iterationCount 1 --launchCount 1`
+    - 结果：通过
+    - 备注：复跑确认后为 `PipelineCount=0` `64.755 ns / 32 B`、`PipelineCount=1` `353.141 ns / 536 B`、`PipelineCount=4` `555.083 ns / 896 B`
+- 本轮结论：
+  - request pipeline benchmark 现在已与默认 request steady-state 使用同一条 generated-provider 宿主接线路径，后续再看 `0 / 1 / 4` 行为矩阵时不再混入“默认 request 已吸收 generated invoker，而 pipeline 还停在纯反射宿主”的口径偏差
+  - `0 pipeline` steady-state 继续下探到约 `64.755 ns / 32 B`，与 `RP-105` 的默认 request benchmark 收敛方向一致，说明这条宿主接线收益能稳定复用到 pipeline benchmark
+  - `1 pipeline` 与 `4 pipeline` 结果在当前 short job 配置下存在噪音，但没有出现清晰的新增分配或显著退化；因此本轮适合作为低风险宿主收口批次接受
+  - 下一批若继续沿用 `$gframework-batch-boot 50`，应优先查看 request lifetime、stream 或 notification benchmark 中是否还存在未吸收 generated-provider 宿主收益的对称切片，而不是回头重试已被 benchmark 否决的 runtime 微优化
+
+### 阶段：默认 request benchmark 吸收 generated provider 宿主（CQRS-REWRITE-RP-105）
+
+- 延续 `$gframework-batch-boot 50`，本轮先确认失败试验已手工回退回 `RP-104` 的已验证状态，再重新评估“默认 request 路径继续逼近 source-generated `Mediator`”的下一刀
+- 本轮接受的只读探索结论：
+  - 继续在 `CqrsDispatcher` 或 `MicrosoftDiContainer` 上堆叠同层级微优化的性价比已经下降，而且上一轮“总是 `GetAll(Type)`”的试验已被 benchmark 明确否决
+  - 默认 `RequestBenchmarks` 虽然已包含 `Mediator` 对照，但当前 GFramework 组仍只注册了单个 handler 实例，没有走 `RegisterCqrsHandlersFromAssembly(...)` + generated registry/provider 的真实宿主接线路径
+  - `RequestInvokerBenchmarks` 已证明 generated request invoker provider 路径比纯反射 binding 更接近目标，因此下一批最小切片应先把这条收益吸收到默认 steady-state request benchmark
+- 本轮主线程决策：
+  - 在 `BenchmarkHostFactory` 内补齐 benchmark 最小宿主的 CQRS 基础设施预接线：runtime、legacy alias、registrar、registration service
+  - 新增 `GeneratedDefaultRequestBenchmarkRegistry`，用 handwritten generated registry + `ICqrsRequestInvokerProvider` + `IEnumeratesCqrsRequestInvokerDescriptors` 为 `RequestBenchmarks.BenchmarkRequest` 提供真实的 generated request invoker descriptor
+  - 让 `RequestBenchmarks` 改用 `RegisterCqrsHandlersFromAssembly(typeof(RequestBenchmarks).Assembly)` 建容器，并在 `Setup/Cleanup` 前后显式清理 dispatcher 静态缓存，避免前一组 benchmark 污染默认 request steady-state 结果
+- 本轮权威验证：
+  - `dotnet build GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.csproj -c Release`
+    - 结果：通过，`0 warning / 0 error`
+  - `python3 scripts/license-header.py --check --paths GFramework.Cqrs.Benchmarks/Messaging/GeneratedDefaultRequestBenchmarkRegistry.cs GFramework.Cqrs.Benchmarks/Messaging/BenchmarkHostFactory.cs GFramework.Cqrs.Benchmarks/Messaging/RequestBenchmarks.cs GFramework.Cqrs.Benchmarks/README.md`
+    - 结果：通过
+  - `git diff --check`
+    - 结果：通过
+  - `dotnet run --project GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.csproj -c Release -- --filter "*RequestBenchmarks.SendRequest_*" --job short --warmupCount 1 --iterationCount 1 --launchCount 1`
+    - 结果：通过
+    - 备注：steady-state request 对照约为 baseline `5.013 ns / 32 B`、`Mediator` `5.747 ns / 32 B`、`MediatR` `51.588 ns / 232 B`、`GFramework.Cqrs` `65.296 ns / 32 B`
+  - `dotnet run --project GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.csproj -c Release -- --filter "*RequestLifetimeBenchmarks.SendRequest_*" --job short --warmupCount 1 --iterationCount 1 --launchCount 1`
+    - 结果：通过
+    - 备注：`Singleton` 下 `GFramework.Cqrs` / `MediatR` 约 `68.772 ns / 32 B` vs `48.177 ns / 232 B`；`Transient` 下约 `73.157 ns / 56 B` vs `51.753 ns / 232 B`
+- 本轮结论：
+  - 默认 request benchmark 现在终于测到了“默认宿主已吸收 generated request invoker provider”后的真实 steady-state，而不再只是纯反射 request binding
+  - 这条宿主层收口在不改 runtime 语义的前提下，把 `GFramework.Cqrs` steady-state request 从约 `70.298 ns` 再压到约 `65.296 ns`
+  - lifetime 矩阵也同步改善到 `68.772 ns / 73.157 ns`，说明默认 request 宿主吸收 generated provider 不只是 benchmark 口径变化，而是对常见 handler 生命周期也有稳定收益
+  - 下一批若继续沿用 `$gframework-batch-boot 50`，应优先转向 pipeline 路径或 handler 解析热路径中仍未吸收 generated/provider 收益的常量开销，而不是回头重试已被否决的 `GetAll(Type)` 零行为探测方案
+
+### 阶段：request 热路径继续收口（CQRS-REWRITE-RP-104）
+
+- 延续 `$gframework-batch-boot 50`，本轮先重新按 `origin/main` 复核 branch diff 基线：
+  - `origin/main` = `4d6dbba6`，提交时间 `2026-05-08 11:13:33 +0800`
+  - 当前分支 `feat/cqrs-optimization` 相对 `origin/main` 的累计 branch diff 仍为 `0 files / 0 lines`
+  - 当前工作树在真正落代码前只有活跃文档更新，仍明显低于 `$gframework-batch-boot 50` 的文件阈值，因此继续自动推进下一批 request 热路径收口
+- 本轮接受的只读探索结论：
+  - `RequestBenchmarks` / `RequestInvokerBenchmarks` 的下一个低风险热点仍在“每次发送都必经的容器查询与短生命周期对象创建”，不是重新回到更高风险的语义层重构
+  - 候选优先级排序为：`SendAsync` 自身状态机开销、`HasRegistration + GetAll` / 服务键扫描，以及 pipeline continuation 的临时对象
+- 本轮主线程决策：
+  - 先以最小行为改动切第一刀：把 `CqrsDispatcher.SendAsync(...)` 从 `async/await` 改为 direct-return `ValueTask`，让零 pipeline request 常见路径不再为 dispatcher 自身生成额外状态机
+  - 在第一刀验证通过且 benchmark 明显改善后，再切第二刀：让 `MicrosoftDiContainer.HasRegistration(Type)` 在冻结后复用预构建的服务键索引，而不是每次线性扫描全部 `ServiceDescriptor`
+  - 第二刀完成后停止继续叠第三刀，因为当前批次已经能清晰区分“有效收益”和“无回退但收益不明显”的因果，不再为了追逐更小常量开销降低评审清晰度
+- 本轮权威验证：
+  - `dotnet build GFramework.Cqrs/GFramework.Cqrs.csproj -c Release`
+    - 结果：通过，`0 warning / 0 error`
+  - `dotnet build GFramework.Core/GFramework.Core.csproj -c Release`
+    - 结果：通过，`0 warning / 0 error`
+  - `dotnet build GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.csproj -c Release`
+    - 结果：通过，`0 warning / 0 error`
+  - `dotnet test GFramework.Core.Tests/GFramework.Core.Tests.csproj -c Release --filter "FullyQualifiedName~MicrosoftDiContainerTests"`
+    - 结果：通过，`52/52` passed
+  - `dotnet test GFramework.Cqrs.Tests/GFramework.Cqrs.Tests.csproj -c Release --no-build --filter "FullyQualifiedName~CqrsDispatcherCacheTests|FullyQualifiedName~CqrsDispatcherContextValidationTests"`
+    - 结果：通过，`14/14` passed
+  - `dotnet run --project GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.csproj -c Release --no-build -- --filter "*RequestBenchmarks.SendRequest_*" --job short --warmupCount 1 --iterationCount 1 --launchCount 1`
+    - 结果：通过
+    - 备注：最新 steady-state request 对照约为 baseline `6.141 ns / 32 B`、`Mediator` `6.674 ns / 32 B`、`MediatR` `61.803 ns / 232 B`、`GFramework.Cqrs` `70.298 ns / 32 B`
+  - `dotnet run --project GFramework.Cqrs.Benchmarks/GFramework.Cqrs.Benchmarks.csproj -c Release --no-build -- --filter "*RequestLifetimeBenchmarks.SendRequest_*" --job short --warmupCount 1 --iterationCount 1 --launchCount 1`
+    - 结果：通过
+    - 备注：最新 lifetime request 对照约为 `Singleton` 下 baseline / `MediatR` / `GFramework.Cqrs` = `4.706 ns / 52.197 ns / 73.005 ns`，`Transient` 下 = `4.571 ns / 50.175 ns / 74.757 ns`
+  - `env GIT_DIR=... GIT_WORK_TREE=... python3 scripts/license-header.py --check`
+    - 结果：通过
+  - `git diff --check`
+    - 结果：通过
+    - 备注：仍仅有 `GFramework.sln` 的历史 CRLF 警告，无本轮新增格式问题
+- 本轮结论：
+  - 第一刀有效：`CqrsDispatcher.SendAsync(...)` 的 direct-return `ValueTask` 把 `GFramework.Cqrs` steady-state request 从 `RP-103` 记录的约 `83.823 ns` 压到约 `70.298 ns`
+  - 第二刀保守有效：冻结后 `HasRegistration(Type)` 索引化没有带来同量级的可见收益，但也没有造成功能回退、额外分配或测试破坏
+  - 下一批若继续压 request hot path，应优先评估默认 request 路径吸收 generated invoker/provider，而不是继续围绕同层级容器存在性判断做微调
+
 ### 阶段：PR #340 latest-head review 收口（CQRS-REWRITE-RP-103）
 
 - 使用 `$gframework-pr-review` 抓取 `feat/cqrs-optimization` 当前公开 PR，并确认当前锚点已从 `PR #339` 更新为 `PR #340`