From f6ffd574bfb2e455d24965441e9254151394ccaf Mon Sep 17 00:00:00 2001
From: MyPrototypeWhat <daoquqiexing@gmail.com>
Date: Mon, 8 Sep 2025 23:29:34 +0800
Subject: [PATCH] refactor(ToolCall): refactor:mcp-tool-state-management
 (#10028)

---
 .../aiCore/examples/hub-provider-usage.ts     | 103 -------
 packages/aiCore/examples/image-generation.ts  | 167 ----------
 packages/aiCore/package.json                  |   6 +-
 .../aiCore/src/core/models/ModelResolver.ts   |   1 -
 .../toolUsePlugin/StreamEventManager.ts       | 101 ++++--
 .../built-in/toolUsePlugin/ToolExecutor.ts    |  57 ++--
 .../toolUsePlugin/promptToolUsePlugin.ts      | 181 +++++++----
 .../plugins/built-in/webSearchPlugin/index.ts |   2 +-
 packages/aiCore/src/core/plugins/index.ts     |   9 +-
 packages/aiCore/src/core/plugins/types.ts     |   2 +-
 .../aiCore/src/core/runtime/pluginEngine.ts   |   6 +-
 .../src/aiCore/chunk/AiSdkToChunkAdapter.ts   |   9 +-
 .../src/aiCore/chunk/handleToolCallChunk.ts   | 289 +++++++++++-------
 src/renderer/src/aiCore/index_new.ts          |  18 +-
 .../src/aiCore/plugins/PluginBuilder.ts       |   7 +-
 .../plugins/searchOrchestrationPlugin.ts      |  43 +--
 .../src/aiCore/tools/MemorySearchTool.ts      |  10 -
 src/renderer/src/aiCore/utils/mcp.ts          |  89 +++---
 .../home/Messages/Blocks/ThinkingBlock.tsx    |   1 -
 .../pages/home/Messages/MessageMenubar.tsx    |   1 -
 .../home/Messages/Tools/MessageMcpTool.tsx    | 135 ++++----
 src/renderer/src/services/ApiService.ts       |   3 +-
 src/renderer/src/services/WebSearchService.ts |   5 -
 .../callbacks/baseCallbacks.ts                |  10 -
 .../callbacks/toolCallbacks.ts                |  24 --
 .../streamCallback.integration.test.ts        |  48 +--
 src/renderer/src/utils/mcp-tools.ts           |   4 +-
 yarn.lock                                     |  51 ++--
 28 files changed, 594 insertions(+), 788 deletions(-)
 delete mode 100644 packages/aiCore/examples/hub-provider-usage.ts
 delete mode 100644 packages/aiCore/examples/image-generation.ts

diff --git a/packages/aiCore/examples/hub-provider-usage.ts b/packages/aiCore/examples/hub-provider-usage.ts
deleted file mode 100644
index 559e812bdb..0000000000
--- a/packages/aiCore/examples/hub-provider-usage.ts
+++ /dev/null
@@ -1,103 +0,0 @@
-/**
- * Hub Provider 使用示例
- *
- * 演示如何使用简化后的Hub Provider功能来路由到多个底层provider
- */
-
-import { createHubProvider, initializeProvider, providerRegistry } from '../src/index'
-
-async function demonstrateHubProvider() {
-  try {
-    // 1. 初始化底层providers
-    console.log('📦 初始化底层providers...')
-
-    initializeProvider('openai', {
-      apiKey: process.env.OPENAI_API_KEY || 'sk-test-key'
-    })
-
-    initializeProvider('anthropic', {
-      apiKey: process.env.ANTHROPIC_API_KEY || 'sk-ant-test-key'
-    })
-
-    // 2. 创建Hub Provider（自动包含所有已初始化的providers）
-    console.log('🌐 创建Hub Provider...')
-
-    const aihubmixProvider = createHubProvider({
-      hubId: 'aihubmix',
-      debug: true
-    })
-
-    // 3. 注册Hub Provider
-    providerRegistry.registerProvider('aihubmix', aihubmixProvider)
-
-    console.log('✅ Hub Provider "aihubmix" 注册成功')
-
-    // 4. 使用Hub Provider访问不同的模型
-    console.log('\n🚀 使用Hub模型...')
-
-    // 通过Hub路由到OpenAI
-    const openaiModel = providerRegistry.languageModel('aihubmix:openai:gpt-4')
-    console.log('✓ OpenAI模型已获取:', openaiModel.modelId)
-
-    // 通过Hub路由到Anthropic
-    const anthropicModel = providerRegistry.languageModel('aihubmix:anthropic:claude-3.5-sonnet')
-    console.log('✓ Anthropic模型已获取:', anthropicModel.modelId)
-
-    // 5. 演示错误处理
-    console.log('\n❌ 演示错误处理...')
-
-    try {
-      // 尝试访问未初始化的provider
-      providerRegistry.languageModel('aihubmix:google:gemini-pro')
-    } catch (error) {
-      console.log('预期错误:', error.message)
-    }
-
-    try {
-      // 尝试使用错误的模型ID格式
-      providerRegistry.languageModel('aihubmix:invalid-format')
-    } catch (error) {
-      console.log('预期错误:', error.message)
-    }
-
-    // 6. 多个Hub Provider示例
-    console.log('\n🔄 创建多个Hub Provider...')
-
-    const localHubProvider = createHubProvider({
-      hubId: 'local-ai'
-    })
-
-    providerRegistry.registerProvider('local-ai', localHubProvider)
-    console.log('✅ Hub Provider "local-ai" 注册成功')
-
-    console.log('\n🎉 Hub Provider演示完成！')
-  } catch (error) {
-    console.error('💥 演示过程中发生错误:', error)
-  }
-}
-
-// 演示简化的使用方式
-function simplifiedUsageExample() {
-  console.log('\n📝 简化使用示例:')
-  console.log(`
-// 1. 初始化providers
-initializeProvider('openai', { apiKey: 'sk-xxx' })
-initializeProvider('anthropic', { apiKey: 'sk-ant-xxx' })
-
-// 2. 创建并注册Hub Provider
-const hubProvider = createHubProvider({ hubId: 'aihubmix' })
-providerRegistry.registerProvider('aihubmix', hubProvider)
-
-// 3. 直接使用
-const model1 = providerRegistry.languageModel('aihubmix:openai:gpt-4')
-const model2 = providerRegistry.languageModel('aihubmix:anthropic:claude-3.5-sonnet')
-`)
-}
-
-// 运行演示
-if (require.main === module) {
-  demonstrateHubProvider()
-  simplifiedUsageExample()
-}
-
-export { demonstrateHubProvider, simplifiedUsageExample }
diff --git a/packages/aiCore/examples/image-generation.ts b/packages/aiCore/examples/image-generation.ts
deleted file mode 100644
index 811aa048c8..0000000000
--- a/packages/aiCore/examples/image-generation.ts
+++ /dev/null
@@ -1,167 +0,0 @@
-/**
- * Image Generation Example
- * 演示如何使用 aiCore 的文生图功能
- */
-
-import { createExecutor, generateImage } from '../src/index'
-
-async function main() {
-  // 方式1: 使用执行器实例
-  console.log('📸 创建 OpenAI 图像生成执行器...')
-  const executor = createExecutor('openai', {
-    apiKey: process.env.OPENAI_API_KEY!
-  })
-
-  try {
-    console.log('🎨 使用执行器生成图像...')
-    const result1 = await executor.generateImage('dall-e-3', {
-      prompt: 'A futuristic cityscape at sunset with flying cars',
-      size: '1024x1024',
-      n: 1
-    })
-
-    console.log('✅ 图像生成成功!')
-    console.log('📊 结果:', {
-      imagesCount: result1.images.length,
-      mediaType: result1.image.mediaType,
-      hasBase64: !!result1.image.base64,
-      providerMetadata: result1.providerMetadata
-    })
-  } catch (error) {
-    console.error('❌ 执行器生成失败:', error)
-  }
-
-  // 方式2: 使用直接调用 API
-  try {
-    console.log('🎨 使用直接 API 生成图像...')
-    const result2 = await generateImage('openai', { apiKey: process.env.OPENAI_API_KEY! }, 'dall-e-3', {
-      prompt: 'A magical forest with glowing mushrooms and fairy lights',
-      aspectRatio: '16:9',
-      providerOptions: {
-        openai: {
-          quality: 'hd',
-          style: 'vivid'
-        }
-      }
-    })
-
-    console.log('✅ 直接 API 生成成功!')
-    console.log('📊 结果:', {
-      imagesCount: result2.images.length,
-      mediaType: result2.image.mediaType,
-      hasBase64: !!result2.image.base64
-    })
-  } catch (error) {
-    console.error('❌ 直接 API 生成失败:', error)
-  }
-
-  // 方式3: 支持其他提供商 (Google Imagen)
-  if (process.env.GOOGLE_API_KEY) {
-    try {
-      console.log('🎨 使用 Google Imagen 生成图像...')
-      const googleExecutor = createExecutor('google', {
-        apiKey: process.env.GOOGLE_API_KEY!
-      })
-
-      const result3 = await googleExecutor.generateImage('imagen-3.0-generate-002', {
-        prompt: 'A serene mountain lake at dawn with mist rising from the water',
-        aspectRatio: '1:1'
-      })
-
-      console.log('✅ Google Imagen 生成成功!')
-      console.log('📊 结果:', {
-        imagesCount: result3.images.length,
-        mediaType: result3.image.mediaType,
-        hasBase64: !!result3.image.base64
-      })
-    } catch (error) {
-      console.error('❌ Google Imagen 生成失败:', error)
-    }
-  }
-
-  // 方式4: 支持插件系统
-  const pluginExample = async () => {
-    console.log('🔌 演示插件系统...')
-
-    // 创建一个示例插件，用于修改提示词
-    const promptEnhancerPlugin = {
-      name: 'prompt-enhancer',
-      transformParams: async (params: any) => {
-        console.log('🔧 插件: 增强提示词...')
-        return {
-          ...params,
-          prompt: `${params.prompt}, highly detailed, cinematic lighting, 4K resolution`
-        }
-      },
-      transformResult: async (result: any) => {
-        console.log('🔧 插件: 处理结果...')
-        return {
-          ...result,
-          enhanced: true
-        }
-      }
-    }
-
-    const executorWithPlugin = createExecutor(
-      'openai',
-      {
-        apiKey: process.env.OPENAI_API_KEY!
-      },
-      [promptEnhancerPlugin]
-    )
-
-    try {
-      const result4 = await executorWithPlugin.generateImage('dall-e-3', {
-        prompt: 'A cute robot playing in a garden'
-      })
-
-      console.log('✅ 插件系统生成成功!')
-      console.log('📊 结果:', {
-        imagesCount: result4.images.length,
-        enhanced: (result4 as any).enhanced,
-        mediaType: result4.image.mediaType
-      })
-    } catch (error) {
-      console.error('❌ 插件系统生成失败:', error)
-    }
-  }
-
-  await pluginExample()
-}
-
-// 错误处理演示
-async function errorHandlingExample() {
-  console.log('⚠️  演示错误处理...')
-
-  try {
-    const executor = createExecutor('openai', {
-      apiKey: 'invalid-key'
-    })
-
-    await executor.generateImage('dall-e-3', {
-      prompt: 'Test image'
-    })
-  } catch (error: any) {
-    console.log('✅ 成功捕获错误:', error.constructor.name)
-    console.log('📋 错误信息:', error.message)
-    console.log('🏷️  提供商ID:', error.providerId)
-    console.log('🏷️  模型ID:', error.modelId)
-  }
-}
-
-// 运行示例
-if (require.main === module) {
-  main()
-    .then(() => {
-      console.log('🎉 所有示例完成!')
-      return errorHandlingExample()
-    })
-    .then(() => {
-      console.log('🎯 示例程序结束')
-      process.exit(0)
-    })
-    .catch((error) => {
-      console.error('💥 程序执行出错:', error)
-      process.exit(1)
-    })
-}
diff --git a/packages/aiCore/package.json b/packages/aiCore/package.json
index 16d9af76a5..950b1d74b2 100644
--- a/packages/aiCore/package.json
+++ b/packages/aiCore/package.json
@@ -1,6 +1,6 @@
 {
   "name": "@cherrystudio/ai-core",
-  "version": "1.0.0-alpha.12",
+  "version": "1.0.0-alpha.13",
   "description": "Cherry Studio AI Core - Unified AI Provider Interface Based on Vercel AI SDK",
   "main": "dist/index.js",
   "module": "dist/index.mjs",
@@ -39,8 +39,8 @@
     "@ai-sdk/anthropic": "^2.0.5",
     "@ai-sdk/azure": "^2.0.16",
     "@ai-sdk/deepseek": "^1.0.9",
-    "@ai-sdk/google": "^2.0.7",
-    "@ai-sdk/openai": "^2.0.19",
+    "@ai-sdk/google": "^2.0.13",
+    "@ai-sdk/openai": "^2.0.26",
     "@ai-sdk/openai-compatible": "^1.0.9",
     "@ai-sdk/provider": "^2.0.0",
     "@ai-sdk/provider-utils": "^3.0.4",
diff --git a/packages/aiCore/src/core/models/ModelResolver.ts b/packages/aiCore/src/core/models/ModelResolver.ts
index 0f1bde95c6..20bf3d76d1 100644
--- a/packages/aiCore/src/core/models/ModelResolver.ts
+++ b/packages/aiCore/src/core/models/ModelResolver.ts
@@ -84,7 +84,6 @@ export class ModelResolver {
    */
   private resolveTraditionalModel(providerId: string, modelId: string): LanguageModelV2 {
     const fullModelId = `${providerId}${DEFAULT_SEPARATOR}${modelId}`
-    console.log('fullModelId', fullModelId)
     return globalRegistryManagement.languageModel(fullModelId as any)
   }
 
diff --git a/packages/aiCore/src/core/plugins/built-in/toolUsePlugin/StreamEventManager.ts b/packages/aiCore/src/core/plugins/built-in/toolUsePlugin/StreamEventManager.ts
index 197b20e9b4..59a425712c 100644
--- a/packages/aiCore/src/core/plugins/built-in/toolUsePlugin/StreamEventManager.ts
+++ b/packages/aiCore/src/core/plugins/built-in/toolUsePlugin/StreamEventManager.ts
@@ -27,10 +27,20 @@ export class StreamEventManager {
   /**
    * 发送步骤完成事件
    */
-  sendStepFinishEvent(controller: StreamController, chunk: any): void {
+  sendStepFinishEvent(
+    controller: StreamController,
+    chunk: any,
+    context: AiRequestContext,
+    finishReason: string = 'stop'
+  ): void {
+    // 累加当前步骤的 usage
+    if (chunk.usage && context.accumulatedUsage) {
+      this.accumulateUsage(context.accumulatedUsage, chunk.usage)
+    }
+
     controller.enqueue({
       type: 'finish-step',
-      finishReason: 'stop',
+      finishReason,
       response: chunk.response,
       usage: chunk.usage,
       providerMetadata: chunk.providerMetadata
@@ -43,28 +53,32 @@ export class StreamEventManager {
   async handleRecursiveCall(
     controller: StreamController,
     recursiveParams: any,
-    context: AiRequestContext,
-    stepId: string
+    context: AiRequestContext
   ): Promise<void> {
-    try {
-      console.log('[MCP Prompt] Starting recursive call after tool execution...')
+    // try {
+    // 重置工具执行状态，准备处理新的步骤
+    context.hasExecutedToolsInCurrentStep = false
 
-      const recursiveResult = await context.recursiveCall(recursiveParams)
+    const recursiveResult = await context.recursiveCall(recursiveParams)
 
-      if (recursiveResult && recursiveResult.fullStream) {
-        await this.pipeRecursiveStream(controller, recursiveResult.fullStream)
-      } else {
-        console.warn('[MCP Prompt] No fullstream found in recursive result:', recursiveResult)
-      }
-    } catch (error) {
-      this.handleRecursiveCallError(controller, error, stepId)
+    if (recursiveResult && recursiveResult.fullStream) {
+      await this.pipeRecursiveStream(controller, recursiveResult.fullStream, context)
+    } else {
+      console.warn('[MCP Prompt] No fullstream found in recursive result:', recursiveResult)
     }
+    // } catch (error) {
+    //   this.handleRecursiveCallError(controller, error, stepId)
+    // }
   }
 
   /**
    * 将递归流的数据传递到当前流
    */
-  private async pipeRecursiveStream(controller: StreamController, recursiveStream: ReadableStream): Promise<void> {
+  private async pipeRecursiveStream(
+    controller: StreamController,
+    recursiveStream: ReadableStream,
+    context?: AiRequestContext
+  ): Promise<void> {
     const reader = recursiveStream.getReader()
     try {
       while (true) {
@@ -73,9 +87,16 @@ export class StreamEventManager {
           break
         }
         if (value.type === 'finish') {
-          // 迭代的流不发finish
+          // 迭代的流不发finish，但需要累加其 usage
+          if (value.usage && context?.accumulatedUsage) {
+            this.accumulateUsage(context.accumulatedUsage, value.usage)
+          }
           break
         }
+        // 对于 finish-step 类型，累加其 usage
+        if (value.type === 'finish-step' && value.usage && context?.accumulatedUsage) {
+          this.accumulateUsage(context.accumulatedUsage, value.usage)
+        }
         // 将递归流的数据传递到当前流
         controller.enqueue(value)
       }
@@ -87,25 +108,25 @@ export class StreamEventManager {
   /**
    * 处理递归调用错误
    */
-  private handleRecursiveCallError(controller: StreamController, error: unknown, stepId: string): void {
-    console.error('[MCP Prompt] Recursive call failed:', error)
+  // private handleRecursiveCallError(controller: StreamController, error: unknown): void {
+  //   console.error('[MCP Prompt] Recursive call failed:', error)
 
-    // 使用 AI SDK 标准错误格式，但不中断流
-    controller.enqueue({
-      type: 'error',
-      error: {
-        message: error instanceof Error ? error.message : String(error),
-        name: error instanceof Error ? error.name : 'RecursiveCallError'
-      }
-    })
+  //   // 使用 AI SDK 标准错误格式，但不中断流
+  //   controller.enqueue({
+  //     type: 'error',
+  //     error: {
+  //       message: error instanceof Error ? error.message : String(error),
+  //       name: error instanceof Error ? error.name : 'RecursiveCallError'
+  //     }
+  //   })
 
-    // 继续发送文本增量，保持流的连续性
-    controller.enqueue({
-      type: 'text-delta',
-      id: stepId,
-      text: '\n\n[工具执行后递归调用失败，继续对话...]'
-    })
-  }
+  //   // // 继续发送文本增量，保持流的连续性
+  //   // controller.enqueue({
+  //   //   type: 'text-delta',
+  //   //   id: stepId,
+  //   //   text: '\n\n[工具执行后递归调用失败，继续对话...]'
+  //   // })
+  // }
 
   /**
    * 构建递归调用的参数
@@ -136,4 +157,18 @@ export class StreamEventManager {
 
     return recursiveParams
   }
+
+  /**
+   * 累加 usage 数据
+   */
+  private accumulateUsage(target: any, source: any): void {
+    if (!target || !source) return
+
+    // 累加各种 token 类型
+    target.inputTokens = (target.inputTokens || 0) + (source.inputTokens || 0)
+    target.outputTokens = (target.outputTokens || 0) + (source.outputTokens || 0)
+    target.totalTokens = (target.totalTokens || 0) + (source.totalTokens || 0)
+    target.reasoningTokens = (target.reasoningTokens || 0) + (source.reasoningTokens || 0)
+    target.cachedInputTokens = (target.cachedInputTokens || 0) + (source.cachedInputTokens || 0)
+  }
 }
diff --git a/packages/aiCore/src/core/plugins/built-in/toolUsePlugin/ToolExecutor.ts b/packages/aiCore/src/core/plugins/built-in/toolUsePlugin/ToolExecutor.ts
index ec174fa2ea..29d644554e 100644
--- a/packages/aiCore/src/core/plugins/built-in/toolUsePlugin/ToolExecutor.ts
+++ b/packages/aiCore/src/core/plugins/built-in/toolUsePlugin/ToolExecutor.ts
@@ -4,7 +4,7 @@
  * 负责工具的执行、结果格式化和相关事件发送
  * 从 promptToolUsePlugin.ts 中提取出来以降低复杂度
  */
-import type { ToolSet } from 'ai'
+import type { ToolSet, TypedToolError } from 'ai'
 
 import type { ToolUseResult } from './type'
 
@@ -38,7 +38,6 @@ export class ToolExecutor {
     controller: StreamController
   ): Promise<ExecutedResult[]> {
     const executedResults: ExecutedResult[] = []
-
     for (const toolUse of toolUses) {
       try {
         const tool = tools[toolUse.toolName]
@@ -46,17 +45,12 @@ export class ToolExecutor {
           throw new Error(`Tool "${toolUse.toolName}" has no execute method`)
         }
 
-        // 发送工具调用开始事件
-        this.sendToolStartEvents(controller, toolUse)
-
-        console.log(`[MCP Prompt Stream] Executing tool: ${toolUse.toolName}`, toolUse.arguments)
-
         // 发送 tool-call 事件
         controller.enqueue({
           type: 'tool-call',
           toolCallId: toolUse.id,
           toolName: toolUse.toolName,
-          input: tool.inputSchema
+          input: toolUse.arguments
         })
 
         const result = await tool.execute(toolUse.arguments, {
@@ -111,45 +105,46 @@ export class ToolExecutor {
   /**
    * 发送工具调用开始相关事件
    */
-  private sendToolStartEvents(controller: StreamController, toolUse: ToolUseResult): void {
-    // 发送 tool-input-start 事件
-    controller.enqueue({
-      type: 'tool-input-start',
-      id: toolUse.id,
-      toolName: toolUse.toolName
-    })
-  }
+  // private sendToolStartEvents(controller: StreamController, toolUse: ToolUseResult): void {
+  //   // 发送 tool-input-start 事件
+  //   controller.enqueue({
+  //     type: 'tool-input-start',
+  //     id: toolUse.id,
+  //     toolName: toolUse.toolName
+  //   })
+  // }
 
   /**
    * 处理工具执行错误
    */
-  private handleToolError(
+  private handleToolError<T extends ToolSet>(
     toolUse: ToolUseResult,
     error: unknown,
     controller: StreamController
-    // _tools: ToolSet
   ): ExecutedResult {
     // 使用 AI SDK 标准错误格式
-    // const toolError: TypedToolError<typeof _tools> = {
-    //   type: 'tool-error',
-    //   toolCallId: toolUse.id,
-    //   toolName: toolUse.toolName,
-    //   input: toolUse.arguments,
-    //   error: error instanceof Error ? error.message : String(error)
-    // }
+    const toolError: TypedToolError<T> = {
+      type: 'tool-error',
+      toolCallId: toolUse.id,
+      toolName: toolUse.toolName,
+      input: toolUse.arguments,
+      error
+    }
 
-    // controller.enqueue(toolError)
+    controller.enqueue(toolError)
 
     // 发送标准错误事件
-    controller.enqueue({
-      type: 'error',
-      error: error instanceof Error ? error.message : String(error)
-    })
+    // controller.enqueue({
+    //   type: 'tool-error',
+    //   toolCallId: toolUse.id,
+    //   error: error instanceof Error ? error.message : String(error),
+    //   input: toolUse.arguments
+    // })
 
     return {
       toolCallId: toolUse.id,
       toolName: toolUse.toolName,
-      result: error instanceof Error ? error.message : String(error),
+      result: error,
       isError: true
     }
   }
diff --git a/packages/aiCore/src/core/plugins/built-in/toolUsePlugin/promptToolUsePlugin.ts b/packages/aiCore/src/core/plugins/built-in/toolUsePlugin/promptToolUsePlugin.ts
index 1d795c94a3..04d3bd5cdb 100644
--- a/packages/aiCore/src/core/plugins/built-in/toolUsePlugin/promptToolUsePlugin.ts
+++ b/packages/aiCore/src/core/plugins/built-in/toolUsePlugin/promptToolUsePlugin.ts
@@ -8,9 +8,19 @@ import type { TextStreamPart, ToolSet } from 'ai'
 import { definePlugin } from '../../index'
 import type { AiRequestContext } from '../../types'
 import { StreamEventManager } from './StreamEventManager'
+import { type TagConfig, TagExtractor } from './tagExtraction'
 import { ToolExecutor } from './ToolExecutor'
 import { PromptToolUseConfig, ToolUseResult } from './type'
 
+/**
+ * 工具使用标签配置
+ */
+const TOOL_USE_TAG_CONFIG: TagConfig = {
+  openingTag: '<tool_use>',
+  closingTag: '</tool_use>',
+  separator: '\n'
+}
+
 /**
  * 默认系统提示符模板（提取自 Cherry Studio）
  */
@@ -249,13 +259,11 @@ export const createPromptToolUsePlugin = (config: PromptToolUseConfig = {}) => {
       }
 
       context.mcpTools = params.tools
-      console.log('tools stored in context', params.tools)
 
       // 构建系统提示符
       const userSystemPrompt = typeof params.system === 'string' ? params.system : ''
       const systemPrompt = buildSystemPrompt(userSystemPrompt, params.tools)
       let systemMessage: string | null = systemPrompt
-      console.log('config.context', context)
       if (config.createSystemMessage) {
         // 🎯 如果用户提供了自定义处理函数，使用它
         systemMessage = config.createSystemMessage(systemPrompt, params, context)
@@ -268,20 +276,40 @@ export const createPromptToolUsePlugin = (config: PromptToolUseConfig = {}) => {
         tools: undefined
       }
       context.originalParams = transformedParams
-      console.log('transformedParams', transformedParams)
       return transformedParams
     },
     transformStream: (_: any, context: AiRequestContext) => () => {
       let textBuffer = ''
-      let stepId = ''
+      // let stepId = ''
 
       if (!context.mcpTools) {
         throw new Error('No tools available')
       }
 
-      // 创建工具执行器和流事件管理器
+      // 从 context 中获取或初始化 usage 累加器
+      if (!context.accumulatedUsage) {
+        context.accumulatedUsage = {
+          inputTokens: 0,
+          outputTokens: 0,
+          totalTokens: 0,
+          reasoningTokens: 0,
+          cachedInputTokens: 0
+        }
+      }
+
+      // 创建工具执行器、流事件管理器和标签提取器
       const toolExecutor = new ToolExecutor()
       const streamEventManager = new StreamEventManager()
+      const tagExtractor = new TagExtractor(TOOL_USE_TAG_CONFIG)
+
+      // 在context中初始化工具执行状态，避免递归调用时状态丢失
+      if (!context.hasExecutedToolsInCurrentStep) {
+        context.hasExecutedToolsInCurrentStep = false
+      }
+
+      // 用于hold text-start事件，直到确认有非工具标签内容
+      let pendingTextStart: TextStreamPart<TOOLS> | null = null
+      let hasStartedText = false
 
       type TOOLS = NonNullable<typeof context.mcpTools>
       return new TransformStream<TextStreamPart<TOOLS>, TextStreamPart<TOOLS>>({
@@ -289,83 +317,106 @@ export const createPromptToolUsePlugin = (config: PromptToolUseConfig = {}) => {
           chunk: TextStreamPart<TOOLS>,
           controller: TransformStreamDefaultController<TextStreamPart<TOOLS>>
         ) {
-          // 收集文本内容
-          if (chunk.type === 'text-delta') {
-            textBuffer += chunk.text || ''
-            stepId = chunk.id || ''
-            controller.enqueue(chunk)
+          // Hold住text-start事件，直到确认有非工具标签内容
+          if ((chunk as any).type === 'text-start') {
+            pendingTextStart = chunk
             return
           }
 
-          if (chunk.type === 'text-end' || chunk.type === 'finish-step') {
-            const tools = context.mcpTools
-            if (!tools || Object.keys(tools).length === 0) {
-              controller.enqueue(chunk)
-              return
-            }
+          // text-delta阶段：收集文本内容并过滤工具标签
+          if (chunk.type === 'text-delta') {
+            textBuffer += chunk.text || ''
+            // stepId = chunk.id || ''
 
-            // 解析工具调用
-            const { results: parsedTools, content: parsedContent } = parseToolUse(textBuffer, tools)
-            const validToolUses = parsedTools.filter((t) => t.status === 'pending')
+            // 使用TagExtractor过滤工具标签，只传递非标签内容到UI层
+            const extractionResults = tagExtractor.processText(chunk.text || '')
 
-            // 如果没有有效的工具调用，直接传递原始事件
-            if (validToolUses.length === 0) {
-              controller.enqueue(chunk)
-              return
-            }
-
-            if (chunk.type === 'text-end') {
-              controller.enqueue({
-                type: 'text-end',
-                id: stepId,
-                providerMetadata: {
-                  text: {
-                    value: parsedContent
-                  }
+            for (const result of extractionResults) {
+              // 只传递非标签内容到UI层
+              if (!result.isTagContent && result.content) {
+                // 如果还没有发送text-start且有pending的text-start，先发送它
+                if (!hasStartedText && pendingTextStart) {
+                  controller.enqueue(pendingTextStart)
+                  hasStartedText = true
+                  pendingTextStart = null
                 }
-              })
-              return
+
+                const filteredChunk = {
+                  ...chunk,
+                  text: result.content
+                }
+                controller.enqueue(filteredChunk)
+              }
+            }
+            return
+          }
+
+          if (chunk.type === 'text-end') {
+            // 只有当已经发送了text-start时才发送text-end
+            if (hasStartedText) {
+              controller.enqueue(chunk)
+            }
+            return
+          }
+
+          if (chunk.type === 'finish-step') {
+            // 统一在finish-step阶段检查并执行工具调用
+            const tools = context.mcpTools
+            if (tools && Object.keys(tools).length > 0 && !context.hasExecutedToolsInCurrentStep) {
+              // 解析完整的textBuffer来检测工具调用
+              const { results: parsedTools } = parseToolUse(textBuffer, tools)
+              const validToolUses = parsedTools.filter((t) => t.status === 'pending')
+
+              if (validToolUses.length > 0) {
+                context.hasExecutedToolsInCurrentStep = true
+
+                // 执行工具调用（不需要手动发送 start-step，外部流已经处理）
+                const executedResults = await toolExecutor.executeTools(validToolUses, tools, controller)
+
+                // 发送步骤完成事件，使用 tool-calls 作为 finishReason
+                streamEventManager.sendStepFinishEvent(controller, chunk, context, 'tool-calls')
+
+                // 处理递归调用
+                const toolResultsText = toolExecutor.formatToolResults(executedResults)
+                const recursiveParams = streamEventManager.buildRecursiveParams(
+                  context,
+                  textBuffer,
+                  toolResultsText,
+                  tools
+                )
+
+                await streamEventManager.handleRecursiveCall(controller, recursiveParams, context)
+                return
+              }
             }
 
-            controller.enqueue({
-              ...chunk,
-              finishReason: 'tool-calls'
-            })
-
-            // 发送步骤开始事件
-            streamEventManager.sendStepStartEvent(controller)
-
-            // 执行工具调用
-            const executedResults = await toolExecutor.executeTools(validToolUses, tools, controller)
-
-            // 发送步骤完成事件
-            streamEventManager.sendStepFinishEvent(controller, chunk)
-
-            // 处理递归调用
-            if (validToolUses.length > 0) {
-              const toolResultsText = toolExecutor.formatToolResults(executedResults)
-              const recursiveParams = streamEventManager.buildRecursiveParams(
-                context,
-                textBuffer,
-                toolResultsText,
-                tools
-              )
-
-              await streamEventManager.handleRecursiveCall(controller, recursiveParams, context, stepId)
-            }
+            // 如果没有执行工具调用，直接传递原始finish-step事件
+            controller.enqueue(chunk)
 
             // 清理状态
             textBuffer = ''
             return
           }
 
-          // 对于其他类型的事件，直接传递
-          controller.enqueue(chunk)
+          // 处理 finish 类型，使用累加后的 totalUsage
+          if (chunk.type === 'finish') {
+            controller.enqueue({
+              ...chunk,
+              totalUsage: context.accumulatedUsage
+            })
+            return
+          }
+
+          // 对于其他类型的事件，直接传递（不包括text-start，已在上面处理）
+          if ((chunk as any).type !== 'text-start') {
+            controller.enqueue(chunk)
+          }
         },
 
         flush() {
-          // 流结束时的清理工作
-          console.log('[MCP Prompt] Stream ended, cleaning up...')
+          // 清理pending状态
+          pendingTextStart = null
+          hasStartedText = false
         }
       })
     }
diff --git a/packages/aiCore/src/core/plugins/built-in/webSearchPlugin/index.ts b/packages/aiCore/src/core/plugins/built-in/webSearchPlugin/index.ts
index 3d549eeac4..d941e0d0c7 100644
--- a/packages/aiCore/src/core/plugins/built-in/webSearchPlugin/index.ts
+++ b/packages/aiCore/src/core/plugins/built-in/webSearchPlugin/index.ts
@@ -27,7 +27,7 @@ export const webSearchPlugin = (config: WebSearchPluginConfig = DEFAULT_WEB_SEAR
         case 'openai': {
           if (config.openai) {
             if (!params.tools) params.tools = {}
-            params.tools.web_search_preview = openai.tools.webSearchPreview(config.openai)
+            params.tools.web_search = openai.tools.webSearch(config.openai)
           }
           break
         }
diff --git a/packages/aiCore/src/core/plugins/index.ts b/packages/aiCore/src/core/plugins/index.ts
index 200188d59d..551b252bf1 100644
--- a/packages/aiCore/src/core/plugins/index.ts
+++ b/packages/aiCore/src/core/plugins/index.ts
@@ -1,5 +1,8 @@
 // 核心类型和接口
 export type { AiPlugin, AiRequestContext, HookResult, PluginManagerConfig } from './types'
+import type { ImageModelV2 } from '@ai-sdk/provider'
+import type { LanguageModel } from 'ai'
+
 import type { ProviderId } from '../providers'
 import type { AiPlugin, AiRequestContext } from './types'
 
@@ -9,16 +12,16 @@ export { PluginManager } from './manager'
 // 工具函数
 export function createContext<T extends ProviderId>(
   providerId: T,
-  modelId: string,
+  model: LanguageModel | ImageModelV2,
   originalParams: any
 ): AiRequestContext {
   return {
     providerId,
-    modelId,
+    model,
     originalParams,
     metadata: {},
     startTime: Date.now(),
-    requestId: `${providerId}-${modelId}-${Date.now()}-${Math.random().toString(36).slice(2)}`,
+    requestId: `${providerId}-${typeof model === 'string' ? model : model?.modelId}-${Date.now()}-${Math.random().toString(36).slice(2)}`,
     // 占位
     recursiveCall: () => Promise.resolve(null)
   }
diff --git a/packages/aiCore/src/core/plugins/types.ts b/packages/aiCore/src/core/plugins/types.ts
index 378fa6c3d3..aaf485ea2f 100644
--- a/packages/aiCore/src/core/plugins/types.ts
+++ b/packages/aiCore/src/core/plugins/types.ts
@@ -14,7 +14,7 @@ export type RecursiveCallFn = (newParams: any) => Promise<any>
  */
 export interface AiRequestContext {
   providerId: ProviderId
-  modelId: string
+  model: LanguageModel | ImageModelV2
   originalParams: any
   metadata: Record<string, any>
   startTime: number
diff --git a/packages/aiCore/src/core/runtime/pluginEngine.ts b/packages/aiCore/src/core/runtime/pluginEngine.ts
index 7a4bb440f7..4d877eb071 100644
--- a/packages/aiCore/src/core/runtime/pluginEngine.ts
+++ b/packages/aiCore/src/core/runtime/pluginEngine.ts
@@ -83,7 +83,7 @@ export class PluginEngine<T extends ProviderId = ProviderId> {
     }
 
     // 使用正确的createContext创建请求上下文
-    const context = _context ? _context : createContext(this.providerId, modelId, params)
+    const context = _context ? _context : createContext(this.providerId, model, params)
 
     // 🔥 为上下文添加递归调用能力
     context.recursiveCall = async (newParams: any): Promise<TResult> => {
@@ -159,7 +159,7 @@ export class PluginEngine<T extends ProviderId = ProviderId> {
     }
 
     // 使用正确的createContext创建请求上下文
-    const context = _context ? _context : createContext(this.providerId, modelId, params)
+    const context = _context ? _context : createContext(this.providerId, model, params)
 
     // 🔥 为上下文添加递归调用能力
     context.recursiveCall = async (newParams: any): Promise<TResult> => {
@@ -235,7 +235,7 @@ export class PluginEngine<T extends ProviderId = ProviderId> {
     }
 
     // 创建请求上下文
-    const context = _context ? _context : createContext(this.providerId, modelId, params)
+    const context = _context ? _context : createContext(this.providerId, model, params)
 
     // 🔥 为上下文添加递归调用能力
     context.recursiveCall = async (newParams: any): Promise<TResult> => {
diff --git a/src/renderer/src/aiCore/chunk/AiSdkToChunkAdapter.ts b/src/renderer/src/aiCore/chunk/AiSdkToChunkAdapter.ts
index a899bcd61b..a28000d9dd 100644
--- a/src/renderer/src/aiCore/chunk/AiSdkToChunkAdapter.ts
+++ b/src/renderer/src/aiCore/chunk/AiSdkToChunkAdapter.ts
@@ -152,12 +152,14 @@ export class AiSdkToChunkAdapter {
       //   this.toolCallHandler.handleToolCallCreated(chunk)
       //   break
       case 'tool-call':
-        // 原始的工具调用（未被中间件处理）
         this.toolCallHandler.handleToolCall(chunk)
         break
 
+      case 'tool-error':
+        this.toolCallHandler.handleToolError(chunk)
+        break
+
       case 'tool-result':
-        // 原始的工具调用结果（未被中间件处理）
         this.toolCallHandler.handleToolResult(chunk)
         break
 
@@ -167,7 +169,6 @@ export class AiSdkToChunkAdapter {
       //     type: ChunkType.LLM_RESPONSE_CREATED
       //   })
       //   break
-      // TODO: 需要区分接口开始和步骤开始
       // case 'start-step':
       //   this.onChunk({
       //     type: ChunkType.BLOCK_CREATED
@@ -305,8 +306,6 @@ export class AiSdkToChunkAdapter {
         break
 
       default:
-      // 其他类型的 chunk 可以忽略或记录日志
-      // console.log('Unhandled AI SDK chunk type:', chunk.type, chunk)
     }
   }
 }
diff --git a/src/renderer/src/aiCore/chunk/handleToolCallChunk.ts b/src/renderer/src/aiCore/chunk/handleToolCallChunk.ts
index 8a24c2d010..6ed89ab5c8 100644
--- a/src/renderer/src/aiCore/chunk/handleToolCallChunk.ts
+++ b/src/renderer/src/aiCore/chunk/handleToolCallChunk.ts
@@ -8,34 +8,61 @@ import { loggerService } from '@logger'
 import { processKnowledgeReferences } from '@renderer/services/KnowledgeService'
 import { BaseTool, MCPTool, MCPToolResponse, NormalToolResponse } from '@renderer/types'
 import { Chunk, ChunkType } from '@renderer/types/chunk'
-import type { ProviderMetadata, ToolSet, TypedToolCall, TypedToolResult } from 'ai'
-// import type {
-//   AnthropicSearchOutput,
-//   WebSearchPluginConfig
-// } from '@cherrystudio/ai-core/core/plugins/built-in/webSearchPlugin'
+import type { ToolSet, TypedToolCall, TypedToolError, TypedToolResult } from 'ai'
 
 const logger = loggerService.withContext('ToolCallChunkHandler')
 
+export type ToolcallsMap = {
+  toolCallId: string
+  toolName: string
+  args: any
+  // mcpTool 现在可以是 MCPTool 或我们为 Provider 工具创建的通用类型
+  tool: BaseTool
+}
 /**
  * 工具调用处理器类
  */
 export class ToolCallChunkHandler {
-  //   private onChunk: (chunk: Chunk) => void
-  private activeToolCalls = new Map<
-    string,
-    {
-      toolCallId: string
-      toolName: string
-      args: any
-      // mcpTool 现在可以是 MCPTool 或我们为 Provider 工具创建的通用类型
-      tool: BaseTool
-    }
-  >()
+  private static globalActiveToolCalls = new Map<string, ToolcallsMap>()
+
+  private activeToolCalls = ToolCallChunkHandler.globalActiveToolCalls
   constructor(
     private onChunk: (chunk: Chunk) => void,
     private mcpTools: MCPTool[]
   ) {}
 
+  /**
+   * 内部静态方法：添加活跃工具调用的核心逻辑
+   */
+  private static addActiveToolCallImpl(toolCallId: string, map: ToolcallsMap): boolean {
+    if (!ToolCallChunkHandler.globalActiveToolCalls.has(toolCallId)) {
+      ToolCallChunkHandler.globalActiveToolCalls.set(toolCallId, map)
+      return true
+    }
+    return false
+  }
+
+  /**
+   * 实例方法：添加活跃工具调用
+   */
+  private addActiveToolCall(toolCallId: string, map: ToolcallsMap): boolean {
+    return ToolCallChunkHandler.addActiveToolCallImpl(toolCallId, map)
+  }
+
+  /**
+   * 获取全局活跃的工具调用
+   */
+  public static getActiveToolCalls() {
+    return ToolCallChunkHandler.globalActiveToolCalls
+  }
+
+  /**
+   * 静态方法：添加活跃工具调用（外部访问）
+   */
+  public static addActiveToolCall(toolCallId: string, map: ToolcallsMap): boolean {
+    return ToolCallChunkHandler.addActiveToolCallImpl(toolCallId, map)
+  }
+
   //   /**
   //    * 设置 onChunk 回调
   //    */
@@ -43,103 +70,103 @@ export class ToolCallChunkHandler {
   //     this.onChunk = callback
   //   }
 
-  handleToolCallCreated(
-    chunk:
-      | {
-          type: 'tool-input-start'
-          id: string
-          toolName: string
-          providerMetadata?: ProviderMetadata
-          providerExecuted?: boolean
-        }
-      | {
-          type: 'tool-input-end'
-          id: string
-          providerMetadata?: ProviderMetadata
-        }
-      | {
-          type: 'tool-input-delta'
-          id: string
-          delta: string
-          providerMetadata?: ProviderMetadata
-        }
-  ): void {
-    switch (chunk.type) {
-      case 'tool-input-start': {
-        // 能拿到说明是mcpTool
-        // if (this.activeToolCalls.get(chunk.id)) return
+  // handleToolCallCreated(
+  //   chunk:
+  //     | {
+  //         type: 'tool-input-start'
+  //         id: string
+  //         toolName: string
+  //         providerMetadata?: ProviderMetadata
+  //         providerExecuted?: boolean
+  //       }
+  //     | {
+  //         type: 'tool-input-end'
+  //         id: string
+  //         providerMetadata?: ProviderMetadata
+  //       }
+  //     | {
+  //         type: 'tool-input-delta'
+  //         id: string
+  //         delta: string
+  //         providerMetadata?: ProviderMetadata
+  //       }
+  // ): void {
+  //   switch (chunk.type) {
+  //     case 'tool-input-start': {
+  //       // 能拿到说明是mcpTool
+  //       // if (this.activeToolCalls.get(chunk.id)) return
 
-        const tool: BaseTool | MCPTool = {
-          id: chunk.id,
-          name: chunk.toolName,
-          description: chunk.toolName,
-          type: chunk.toolName.startsWith('builtin_') ? 'builtin' : 'provider'
-        }
-        this.activeToolCalls.set(chunk.id, {
-          toolCallId: chunk.id,
-          toolName: chunk.toolName,
-          args: '',
-          tool
-        })
-        const toolResponse: MCPToolResponse | NormalToolResponse = {
-          id: chunk.id,
-          tool: tool,
-          arguments: {},
-          status: 'pending',
-          toolCallId: chunk.id
-        }
-        this.onChunk({
-          type: ChunkType.MCP_TOOL_PENDING,
-          responses: [toolResponse]
-        })
-        break
-      }
-      case 'tool-input-delta': {
-        const toolCall = this.activeToolCalls.get(chunk.id)
-        if (!toolCall) {
-          logger.warn(`🔧 [ToolCallChunkHandler] Tool call not found: ${chunk.id}`)
-          return
-        }
-        toolCall.args += chunk.delta
-        break
-      }
-      case 'tool-input-end': {
-        const toolCall = this.activeToolCalls.get(chunk.id)
-        this.activeToolCalls.delete(chunk.id)
-        if (!toolCall) {
-          logger.warn(`🔧 [ToolCallChunkHandler] Tool call not found: ${chunk.id}`)
-          return
-        }
-        // const toolResponse: ToolCallResponse = {
-        //   id: toolCall.toolCallId,
-        //   tool: toolCall.tool,
-        //   arguments: toolCall.args,
-        //   status: 'pending',
-        //   toolCallId: toolCall.toolCallId
-        // }
-        // logger.debug('toolResponse', toolResponse)
-        // this.onChunk({
-        //   type: ChunkType.MCP_TOOL_PENDING,
-        //   responses: [toolResponse]
-        // })
-        break
-      }
-    }
-    // if (!toolCall) {
-    //   Logger.warn(`🔧 [ToolCallChunkHandler] Tool call not found: ${chunk.id}`)
-    //   return
-    // }
-    // this.onChunk({
-    //   type: ChunkType.MCP_TOOL_CREATED,
-    //   tool_calls: [
-    //     {
-    //       id: chunk.id,
-    //       name: chunk.toolName,
-    //       status: 'pending'
-    //     }
-    //   ]
-    // })
-  }
+  //       const tool: BaseTool | MCPTool = {
+  //         id: chunk.id,
+  //         name: chunk.toolName,
+  //         description: chunk.toolName,
+  //         type: chunk.toolName.startsWith('builtin_') ? 'builtin' : 'provider'
+  //       }
+  //       this.activeToolCalls.set(chunk.id, {
+  //         toolCallId: chunk.id,
+  //         toolName: chunk.toolName,
+  //         args: '',
+  //         tool
+  //       })
+  //       const toolResponse: MCPToolResponse | NormalToolResponse = {
+  //         id: chunk.id,
+  //         tool: tool,
+  //         arguments: {},
+  //         status: 'pending',
+  //         toolCallId: chunk.id
+  //       }
+  //       this.onChunk({
+  //         type: ChunkType.MCP_TOOL_PENDING,
+  //         responses: [toolResponse]
+  //       })
+  //       break
+  //     }
+  //     case 'tool-input-delta': {
+  //       const toolCall = this.activeToolCalls.get(chunk.id)
+  //       if (!toolCall) {
+  //         logger.warn(`🔧 [ToolCallChunkHandler] Tool call not found: ${chunk.id}`)
+  //         return
+  //       }
+  //       toolCall.args += chunk.delta
+  //       break
+  //     }
+  //     case 'tool-input-end': {
+  //       const toolCall = this.activeToolCalls.get(chunk.id)
+  //       this.activeToolCalls.delete(chunk.id)
+  //       if (!toolCall) {
+  //         logger.warn(`🔧 [ToolCallChunkHandler] Tool call not found: ${chunk.id}`)
+  //         return
+  //       }
+  //       // const toolResponse: ToolCallResponse = {
+  //       //   id: toolCall.toolCallId,
+  //       //   tool: toolCall.tool,
+  //       //   arguments: toolCall.args,
+  //       //   status: 'pending',
+  //       //   toolCallId: toolCall.toolCallId
+  //       // }
+  //       // logger.debug('toolResponse', toolResponse)
+  //       // this.onChunk({
+  //       //   type: ChunkType.MCP_TOOL_PENDING,
+  //       //   responses: [toolResponse]
+  //       // })
+  //       break
+  //     }
+  //   }
+  //   // if (!toolCall) {
+  //   //   Logger.warn(`🔧 [ToolCallChunkHandler] Tool call not found: ${chunk.id}`)
+  //   //   return
+  //   // }
+  //   // this.onChunk({
+  //   //   type: ChunkType.MCP_TOOL_CREATED,
+  //   //   tool_calls: [
+  //   //     {
+  //   //       id: chunk.id,
+  //   //       name: chunk.toolName,
+  //   //       status: 'pending'
+  //   //     }
+  //   //   ]
+  //   // })
+  // }
 
   /**
    * 处理工具调用事件
@@ -158,7 +185,6 @@ export class ToolCallChunkHandler {
 
     let tool: BaseTool
     let mcpTool: MCPTool | undefined
-
     // 根据 providerExecuted 标志区分处理逻辑
     if (providerExecuted) {
       // 如果是 Provider 执行的工具（如 web_search）
@@ -196,27 +222,25 @@ export class ToolCallChunkHandler {
       }
     }
 
-    // 记录活跃的工具调用
-    this.activeToolCalls.set(toolCallId, {
+    this.addActiveToolCall(toolCallId, {
       toolCallId,
       toolName,
       args,
       tool
     })
-
     // 创建 MCPToolResponse 格式
     const toolResponse: MCPToolResponse | NormalToolResponse = {
       id: toolCallId,
       tool: tool,
       arguments: args,
-      status: 'pending',
+      status: 'pending', // 统一使用 pending 状态
       toolCallId: toolCallId
     }
 
     // 调用 onChunk
     if (this.onChunk) {
       this.onChunk({
-        type: ChunkType.MCP_TOOL_PENDING,
+        type: ChunkType.MCP_TOOL_PENDING, // 统一发送 pending 状态
         responses: [toolResponse]
       })
     }
@@ -276,4 +300,33 @@ export class ToolCallChunkHandler {
       })
     }
   }
+  handleToolError(
+    chunk: {
+      type: 'tool-error'
+    } & TypedToolError<ToolSet>
+  ): void {
+    const { toolCallId, error, input } = chunk
+    const toolCallInfo = this.activeToolCalls.get(toolCallId)
+    if (!toolCallInfo) {
+      logger.warn(`🔧 [ToolCallChunkHandler] Tool call info not found for ID: ${toolCallId}`)
+      return
+    }
+    const toolResponse: MCPToolResponse | NormalToolResponse = {
+      id: toolCallId,
+      tool: toolCallInfo.tool,
+      arguments: input,
+      status: 'error',
+      response: error,
+      toolCallId: toolCallId
+    }
+    this.activeToolCalls.delete(toolCallId)
+    if (this.onChunk) {
+      this.onChunk({
+        type: ChunkType.MCP_TOOL_COMPLETE,
+        responses: [toolResponse]
+      })
+    }
+  }
 }
+
+export const addActiveToolCall = ToolCallChunkHandler.addActiveToolCall.bind(ToolCallChunkHandler)
diff --git a/src/renderer/src/aiCore/index_new.ts b/src/renderer/src/aiCore/index_new.ts
index 897fc1c48a..8da8e8c7c1 100644
--- a/src/renderer/src/aiCore/index_new.ts
+++ b/src/renderer/src/aiCore/index_new.ts
@@ -265,15 +265,15 @@ export default class ModernAiProvider {
     params: StreamTextParams,
     config: ModernAiProviderConfig
   ): Promise<CompletionsResult> {
-    const modelId = this.model!.id
-    logger.info('Starting modernCompletions', {
-      modelId,
-      providerId: this.config!.providerId,
-      topicId: config.topicId,
-      hasOnChunk: !!config.onChunk,
-      hasTools: !!params.tools && Object.keys(params.tools).length > 0,
-      toolCount: params.tools ? Object.keys(params.tools).length : 0
-    })
+    // const modelId = this.model!.id
+    // logger.info('Starting modernCompletions', {
+    //   modelId,
+    //   providerId: this.config!.providerId,
+    //   topicId: config.topicId,
+    //   hasOnChunk: !!config.onChunk,
+    //   hasTools: !!params.tools && Object.keys(params.tools).length > 0,
+    //   toolCount: params.tools ? Object.keys(params.tools).length : 0
+    // })
 
     // 根据条件构建插件数组
     const plugins = buildPlugins(config)
diff --git a/src/renderer/src/aiCore/plugins/PluginBuilder.ts b/src/renderer/src/aiCore/plugins/PluginBuilder.ts
index 0f5d7ecc4b..16cf56244d 100644
--- a/src/renderer/src/aiCore/plugins/PluginBuilder.ts
+++ b/src/renderer/src/aiCore/plugins/PluginBuilder.ts
@@ -35,7 +35,7 @@ export function buildPlugins(
     plugins.push(webSearchPlugin())
   }
   // 2. 支持工具调用时添加搜索插件
-  if (middlewareConfig.isSupportedToolUse) {
+  if (middlewareConfig.isSupportedToolUse || middlewareConfig.isPromptToolUse) {
     plugins.push(searchOrchestrationPlugin(middlewareConfig.assistant, middlewareConfig.topicId || ''))
   }
 
@@ -45,12 +45,13 @@ export function buildPlugins(
   }
 
   // 4. 启用Prompt工具调用时添加工具插件
-  if (middlewareConfig.isPromptToolUse && middlewareConfig.mcpTools && middlewareConfig.mcpTools.length > 0) {
+  if (middlewareConfig.isPromptToolUse) {
     plugins.push(
       createPromptToolUsePlugin({
         enabled: true,
         createSystemMessage: (systemPrompt, params, context) => {
-          if (context.modelId.includes('o1-mini') || context.modelId.includes('o1-preview')) {
+          const modelId = typeof context.model === 'string' ? context.model : context.model.modelId
+          if (modelId.includes('o1-mini') || modelId.includes('o1-preview')) {
             if (context.isRecursiveCall) {
               return null
             }
diff --git a/src/renderer/src/aiCore/plugins/searchOrchestrationPlugin.ts b/src/renderer/src/aiCore/plugins/searchOrchestrationPlugin.ts
index 3e6ca8cc3d..afbaf936ef 100644
--- a/src/renderer/src/aiCore/plugins/searchOrchestrationPlugin.ts
+++ b/src/renderer/src/aiCore/plugins/searchOrchestrationPlugin.ts
@@ -19,7 +19,8 @@ import store from '@renderer/store'
 import { selectCurrentUserId, selectGlobalMemoryEnabled, selectMemoryConfig } from '@renderer/store/memory'
 import type { Assistant } from '@renderer/types'
 import { extractInfoFromXML, ExtractResults } from '@renderer/utils/extract'
-import type { ModelMessage } from 'ai'
+import type { LanguageModel, ModelMessage } from 'ai'
+import { generateText } from 'ai'
 import { isEmpty } from 'lodash'
 
 import { MemoryProcessor } from '../../services/MemoryProcessor'
@@ -76,9 +77,7 @@ async function analyzeSearchIntent(
     shouldKnowledgeSearch?: boolean
     shouldMemorySearch?: boolean
     lastAnswer?: ModelMessage
-    context: AiRequestContext & {
-      isAnalyzing?: boolean
-    }
+    context: AiRequestContext
     topicId: string
   }
 ): Promise<ExtractResults | undefined> {
@@ -122,9 +121,7 @@ async function analyzeSearchIntent(
     logger.error('Provider not found or missing API key')
     return getFallbackResult()
   }
-  // console.log('formattedPrompt', schema)
   try {
-    context.isAnalyzing = true
     logger.info('Starting intent analysis generateText call', {
       modelId: model.id,
       topicId: options.topicId,
@@ -133,18 +130,16 @@ async function analyzeSearchIntent(
       hasKnowledgeSearch: needKnowledgeExtract
     })
 
-    const { text: result } = await context.executor
-      .generateText(model.id, {
-        prompt: formattedPrompt
-      })
-      .finally(() => {
-        context.isAnalyzing = false
-        logger.info('Intent analysis generateText call completed', {
-          modelId: model.id,
-          topicId: options.topicId,
-          requestId: context.requestId
-        })
+    const { text: result } = await generateText({
+      model: context.model as LanguageModel,
+      prompt: formattedPrompt
+    }).finally(() => {
+      logger.info('Intent analysis generateText call completed', {
+        modelId: model.id,
+        topicId: options.topicId,
+        requestId: context.requestId
       })
+    })
     const parsedResult = extractInfoFromXML(result)
     logger.debug('Intent analysis result', { parsedResult })
 
@@ -183,7 +178,6 @@ async function storeConversationMemory(
   const globalMemoryEnabled = selectGlobalMemoryEnabled(store.getState())
 
   if (!globalMemoryEnabled || !assistant.enableMemory) {
-    // console.log('Memory storage is disabled')
     return
   }
 
@@ -245,25 +239,14 @@ export const searchOrchestrationPlugin = (assistant: Assistant, topicId: string)
   // 存储意图分析结果
   const intentAnalysisResults: { [requestId: string]: ExtractResults } = {}
   const userMessages: { [requestId: string]: ModelMessage } = {}
-  let currentContext: AiRequestContext | null = null
 
   return definePlugin({
     name: 'search-orchestration',
     enforce: 'pre', // 确保在其他插件之前执行
-
-    configureContext: (context: AiRequestContext) => {
-      if (currentContext) {
-        context.isAnalyzing = currentContext.isAnalyzing
-      }
-      currentContext = context
-    },
-
     /**
      * 🔍 Step 1: 意图识别阶段
      */
     onRequestStart: async (context: AiRequestContext) => {
-      if (context.isAnalyzing) return
-
       // 没开启任何搜索则不进行意图分析
       if (!(assistant.webSearchProviderId || assistant.knowledge_bases?.length || assistant.enableMemory)) return
 
@@ -315,7 +298,6 @@ export const searchOrchestrationPlugin = (assistant: Assistant, topicId: string)
      * 🔧 Step 2: 工具配置阶段
      */
     transformParams: async (params: any, context: AiRequestContext) => {
-      if (context.isAnalyzing) return params
       // logger.info('🔧 Configuring tools based on intent...', context.requestId)
 
       try {
@@ -409,7 +391,6 @@ export const searchOrchestrationPlugin = (assistant: Assistant, topicId: string)
       // context.isAnalyzing = false
       // logger.info('context.isAnalyzing', context, result)
       // logger.info('💾 Starting memory storage...', context.requestId)
-      if (context.isAnalyzing) return
       try {
         const messages = context.originalParams.messages
 
diff --git a/src/renderer/src/aiCore/tools/MemorySearchTool.ts b/src/renderer/src/aiCore/tools/MemorySearchTool.ts
index d8e402db5d..ac70140a59 100644
--- a/src/renderer/src/aiCore/tools/MemorySearchTool.ts
+++ b/src/renderer/src/aiCore/tools/MemorySearchTool.ts
@@ -19,8 +19,6 @@ export const memorySearchTool = () => {
       limit: z.number().min(1).max(20).default(5).describe('Maximum number of memories to return')
     }),
     execute: async ({ query, limit = 5 }) => {
-      // console.log('🧠 [memorySearchTool] Searching memories:', { query, limit })
-
       try {
         const globalMemoryEnabled = selectGlobalMemoryEnabled(store.getState())
         if (!globalMemoryEnabled) {
@@ -29,7 +27,6 @@ export const memorySearchTool = () => {
 
         const memoryConfig = selectMemoryConfig(store.getState())
         if (!memoryConfig.llmApiClient || !memoryConfig.embedderApiClient) {
-          // console.warn('Memory search skipped: embedding or LLM model not configured')
           return []
         }
 
@@ -40,12 +37,10 @@ export const memorySearchTool = () => {
         const relevantMemories = await memoryProcessor.searchRelevantMemories(query, processorConfig, limit)
 
         if (relevantMemories?.length > 0) {
-          // console.log('🧠 [memorySearchTool] Found memories:', relevantMemories.length)
           return relevantMemories
         }
         return []
       } catch (error) {
-        // console.error('🧠 [memorySearchTool] Error:', error)
         return []
       }
     }
@@ -84,8 +79,6 @@ export const memorySearchToolWithExtraction = (assistant: Assistant) => {
         .optional()
     }) satisfies z.ZodSchema<MemorySearchWithExtractionInput>,
     execute: async ({ userMessage }) => {
-      // console.log('🧠 [memorySearchToolWithExtraction] Processing:', { userMessage, lastAnswer })
-
       try {
         const globalMemoryEnabled = selectGlobalMemoryEnabled(store.getState())
         if (!globalMemoryEnabled || !assistant.enableMemory) {
@@ -97,7 +90,6 @@ export const memorySearchToolWithExtraction = (assistant: Assistant) => {
 
         const memoryConfig = selectMemoryConfig(store.getState())
         if (!memoryConfig.llmApiClient || !memoryConfig.embedderApiClient) {
-          // console.warn('Memory search skipped: embedding or LLM model not configured')
           return {
             extractedKeywords: 'Memory models not configured',
             searchResults: []
@@ -125,7 +117,6 @@ export const memorySearchToolWithExtraction = (assistant: Assistant) => {
         )
 
         if (relevantMemories?.length > 0) {
-          // console.log('🧠 [memorySearchToolWithExtraction] Found memories:', relevantMemories.length)
           return {
             extractedKeywords: content,
             searchResults: relevantMemories
@@ -137,7 +128,6 @@ export const memorySearchToolWithExtraction = (assistant: Assistant) => {
           searchResults: []
         }
       } catch (error) {
-        // console.error('🧠 [memorySearchToolWithExtraction] Error:', error)
         return {
           extractedKeywords: 'Search failed',
           searchResults: []
diff --git a/src/renderer/src/aiCore/utils/mcp.ts b/src/renderer/src/aiCore/utils/mcp.ts
index 01f7adbf50..e7f6b5a393 100644
--- a/src/renderer/src/aiCore/utils/mcp.ts
+++ b/src/renderer/src/aiCore/utils/mcp.ts
@@ -1,7 +1,5 @@
 import { loggerService } from '@logger'
-// import { AiSdkTool, ToolCallResult } from '@renderer/aiCore/tools/types'
 import { MCPTool, MCPToolResponse } from '@renderer/types'
-import { Chunk, ChunkType } from '@renderer/types/chunk'
 import { callMCPTool, getMcpServerByTool, isToolAutoApproved } from '@renderer/utils/mcp-tools'
 import { requestToolConfirmation } from '@renderer/utils/userConfirmation'
 import { type Tool, type ToolSet } from 'ai'
@@ -33,8 +31,36 @@ export function convertMcpToolsToAiSdkTools(mcpTools: MCPTool[]): ToolSet {
     tools[mcpTool.name] = tool({
       description: mcpTool.description || `Tool from ${mcpTool.serverName}`,
       inputSchema: jsonSchema(mcpTool.inputSchema as JSONSchema7),
-      execute: async (params, { toolCallId, experimental_context }) => {
-        const { onChunk } = experimental_context as { onChunk: (chunk: Chunk) => void }
+      execute: async (params, { toolCallId }) => {
+        // 检查是否启用自动批准
+        const server = getMcpServerByTool(mcpTool)
+        const isAutoApproveEnabled = isToolAutoApproved(mcpTool, server)
+
+        let confirmed = true
+
+        if (!isAutoApproveEnabled) {
+          // 请求用户确认
+          logger.debug(`Requesting user confirmation for tool: ${mcpTool.name}`)
+          confirmed = await requestToolConfirmation(toolCallId)
+        }
+
+        if (!confirmed) {
+          // 用户拒绝执行工具
+          logger.debug(`User cancelled tool execution: ${mcpTool.name}`)
+          return {
+            content: [
+              {
+                type: 'text',
+                text: `User declined to execute tool "${mcpTool.name}".`
+              }
+            ],
+            isError: false
+          }
+        }
+
+        // 用户确认或自动批准，执行工具
+        logger.debug(`Executing tool: ${mcpTool.name}`)
+
         // 创建适配的 MCPToolResponse 对象
         const toolResponse: MCPToolResponse = {
           id: toolCallId,
@@ -44,53 +70,18 @@ export function convertMcpToolsToAiSdkTools(mcpTools: MCPTool[]): ToolSet {
           toolCallId
         }
 
-        try {
-          // 检查是否启用自动批准
-          const server = getMcpServerByTool(mcpTool)
-          const isAutoApproveEnabled = isToolAutoApproved(mcpTool, server)
+        const result = await callMCPTool(toolResponse)
 
-          let confirmed = true
-          if (!isAutoApproveEnabled) {
-            // 请求用户确认
-            logger.debug(`Requesting user confirmation for tool: ${mcpTool.name}`)
-            confirmed = await requestToolConfirmation(toolResponse.id)
-          }
-
-          if (!confirmed) {
-            // 用户拒绝执行工具
-            logger.debug(`User cancelled tool execution: ${mcpTool.name}`)
-            return {
-              content: [
-                {
-                  type: 'text',
-                  text: `User declined to execute tool "${mcpTool.name}".`
-                }
-              ],
-              isError: false
-            }
-          }
-
-          // 用户确认或自动批准，执行工具
-          toolResponse.status = 'invoking'
-          logger.debug(`Executing tool: ${mcpTool.name}`)
-
-          onChunk({
-            type: ChunkType.MCP_TOOL_IN_PROGRESS,
-            responses: [toolResponse]
-          })
-
-          const result = await callMCPTool(toolResponse)
-
-          // 返回结果，AI SDK 会处理序列化
-          if (result.isError) {
-            throw new Error(result.content?.[0]?.text || 'Tool execution failed')
-          }
-          // 返回工具执行结果
-          return result
-        } catch (error) {
-          logger.error(`MCP Tool execution failed: ${mcpTool.name}`, { error })
-          throw error
+        // 返回结果，AI SDK 会处理序列化
+        if (result.isError) {
+          // throw new Error(result.content?.[0]?.text || 'Tool execution failed')
+          return Promise.reject(result)
         }
+        // 返回工具执行结果
+        return result
+        // } catch (error) {
+        //   logger.error(`MCP Tool execution failed: ${mcpTool.name}`, { error })
+        // }
       }
     })
   }
diff --git a/src/renderer/src/pages/home/Messages/Blocks/ThinkingBlock.tsx b/src/renderer/src/pages/home/Messages/Blocks/ThinkingBlock.tsx
index 0aee1eccac..98cad2a8ca 100644
--- a/src/renderer/src/pages/home/Messages/Blocks/ThinkingBlock.tsx
+++ b/src/renderer/src/pages/home/Messages/Blocks/ThinkingBlock.tsx
@@ -105,7 +105,6 @@ const ThinkingBlock: React.FC<Props> = ({ block }) => {
 const ThinkingTimeSeconds = memo(
   ({ blockThinkingTime, isThinking }: { blockThinkingTime: number; isThinking: boolean }) => {
     const { t } = useTranslation()
-    // console.log('blockThinkingTime', blockThinkingTime)
     // const [thinkingTime, setThinkingTime] = useState(blockThinkingTime || 0)
 
     // FIXME: 这里统计的和请求处统计的有一定误差
diff --git a/src/renderer/src/pages/home/Messages/MessageMenubar.tsx b/src/renderer/src/pages/home/Messages/MessageMenubar.tsx
index 76ad9a8750..0b0c623d28 100644
--- a/src/renderer/src/pages/home/Messages/MessageMenubar.tsx
+++ b/src/renderer/src/pages/home/Messages/MessageMenubar.tsx
@@ -186,7 +186,6 @@ const MessageMenubar: FC<Props> = (props) => {
       try {
         await translateText(mainTextContent, language, translationUpdater)
       } catch (error) {
-        // console.error('Translation failed:', error)
         window.message.error({ content: t('translate.error.failed'), key: 'translate-message' })
         // 理应只有一个
         const translationBlocks = findTranslationBlocksById(message.id)
diff --git a/src/renderer/src/pages/home/Messages/Tools/MessageMcpTool.tsx b/src/renderer/src/pages/home/Messages/Tools/MessageMcpTool.tsx
index 3c56e2f1d7..d3fddd0a1b 100644
--- a/src/renderer/src/pages/home/Messages/Tools/MessageMcpTool.tsx
+++ b/src/renderer/src/pages/home/Messages/Tools/MessageMcpTool.tsx
@@ -60,14 +60,29 @@ const MessageMcpTool: FC<Props> = ({ block }) => {
   const toolResponse = block.metadata?.rawMcpToolResponse
 
   const { id, tool, status, response } = toolResponse!
-
   const isPending = status === 'pending'
-  const isInvoking = status === 'invoking'
   const isDone = status === 'done'
+  const isError = status === 'error'
+
+  const isAutoApproved = useMemo(
+    () =>
+      isToolAutoApproved(
+        tool,
+        mcpServers.find((s) => s.id === tool.serverId)
+      ),
+    [tool, mcpServers]
+  )
+
+  // 增加本地状态来跟踪用户确认
+  const [isConfirmed, setIsConfirmed] = useState(isAutoApproved)
+
+  // 判断不同的UI状态
+  const isWaitingConfirmation = isPending && !isAutoApproved && !isConfirmed
+  const isExecuting = isPending && (isAutoApproved || isConfirmed)
 
   const timer = useRef<NodeJS.Timeout | null>(null)
   useEffect(() => {
-    if (!isPending) return
+    if (!isWaitingConfirmation) return
 
     if (countdown > 0) {
       timer.current = setTimeout(() => {
@@ -75,6 +90,7 @@ const MessageMcpTool: FC<Props> = ({ block }) => {
         setCountdown((prev) => prev - 1)
       }, 1000)
     } else if (countdown === 0) {
+      setIsConfirmed(true)
       confirmToolAction(id)
     }
 
@@ -83,7 +99,7 @@ const MessageMcpTool: FC<Props> = ({ block }) => {
         clearTimeout(timer.current)
       }
     }
-  }, [countdown, id, isPending])
+  }, [countdown, id, isWaitingConfirmation])
 
   useEffect(() => {
     const removeListener = window.electron.ipcRenderer.on(
@@ -146,6 +162,7 @@ const MessageMcpTool: FC<Props> = ({ block }) => {
 
   const handleConfirmTool = () => {
     cancelCountdown()
+    setIsConfirmed(true)
     confirmToolAction(id)
   }
 
@@ -195,6 +212,7 @@ const MessageMcpTool: FC<Props> = ({ block }) => {
     updateMCPServer(updatedServer)
 
     // Also confirm the current tool
+    setIsConfirmed(true)
     confirmToolAction(id)
 
     window.message.success({
@@ -206,32 +224,31 @@ const MessageMcpTool: FC<Props> = ({ block }) => {
   const renderStatusIndicator = (status: string, hasError: boolean) => {
     let label = ''
     let icon: React.ReactNode | null = null
-    switch (status) {
-      case 'pending':
+
+    if (status === 'pending') {
+      if (isWaitingConfirmation) {
         label = t('message.tools.pending', 'Awaiting Approval')
         icon = <LoadingIcon style={{ marginLeft: 6, color: 'var(--status-color-warning)' }} />
-        break
-      case 'invoking':
+      } else if (isExecuting) {
         label = t('message.tools.invoking')
         icon = <LoadingIcon style={{ marginLeft: 6 }} />
-        break
-      case 'cancelled':
-        label = t('message.tools.cancelled')
-        icon = <X size={13} style={{ marginLeft: 6 }} className="lucide-custom" />
-        break
-      case 'done':
-        if (hasError) {
-          label = t('message.tools.error')
-          icon = <TriangleAlert size={13} style={{ marginLeft: 6 }} className="lucide-custom" />
-        } else {
-          label = t('message.tools.completed')
-          icon = <Check size={13} style={{ marginLeft: 6 }} className="lucide-custom" />
-        }
-        break
-      default:
-        label = ''
-        icon = null
+      }
+    } else if (status === 'cancelled') {
+      label = t('message.tools.cancelled')
+      icon = <X size={13} style={{ marginLeft: 6 }} className="lucide-custom" />
+    } else if (status === 'done') {
+      if (hasError) {
+        label = t('message.tools.error')
+        icon = <TriangleAlert size={13} style={{ marginLeft: 6 }} className="lucide-custom" />
+      } else {
+        label = t('message.tools.completed')
+        icon = <Check size={13} style={{ marginLeft: 6 }} className="lucide-custom" />
+      }
+    } else if (status === 'error') {
+      label = t('message.tools.error')
+      icon = <TriangleAlert size={13} style={{ marginLeft: 6 }} className="lucide-custom" />
     }
+
     return (
       <StatusIndicator status={status} hasError={hasError}>
         {label}
@@ -248,7 +265,6 @@ const MessageMcpTool: FC<Props> = ({ block }) => {
       params: toolResponse.arguments,
       response: toolResponse.response
     }
-
     items.push({
       key: id,
       label: (
@@ -283,7 +299,7 @@ const MessageMcpTool: FC<Props> = ({ block }) => {
                 <Maximize size={14} />
               </ActionButton>
             </Tooltip>
-            {!isPending && !isInvoking && (
+            {!isPending && (
               <Tooltip title={t('common.copy')} mouseEnterDelay={0.5}>
                 <ActionButton
                   className="message-action-button"
@@ -301,7 +317,7 @@ const MessageMcpTool: FC<Props> = ({ block }) => {
         </MessageTitleLabel>
       ),
       children:
-        isDone && result ? (
+        (isDone || isError) && result ? (
           <ToolResponseContainer
             style={{
               fontFamily: messageFont === 'serif' ? 'var(--font-family-serif)' : 'var(--font-family)',
@@ -370,7 +386,7 @@ const MessageMcpTool: FC<Props> = ({ block }) => {
           }
         }}>
         <ToolContainer>
-          <ToolContentWrapper className={status}>
+          <ToolContentWrapper className={isPending ? 'pending' : status}>
             <CollapseContainer
               ghost
               activeKey={activeKeys}
@@ -383,14 +399,16 @@ const MessageMcpTool: FC<Props> = ({ block }) => {
                 <ExpandIcon $isActive={isActive} size={18} color="var(--color-text-3)" strokeWidth={1.5} />
               )}
             />
-            {(isPending || isInvoking) && (
+            {isPending && (
               <ActionsBar>
                 <ActionLabel>
-                  {isPending ? t('settings.mcp.tools.autoApprove.tooltip.confirm') : t('message.tools.invoking')}
+                  {isWaitingConfirmation
+                    ? t('settings.mcp.tools.autoApprove.tooltip.confirm')
+                    : t('message.tools.invoking')}
                 </ActionLabel>
 
                 <ActionButtonsGroup>
-                  {isPending && (
+                  {isWaitingConfirmation && (
                     <Button
                       color="danger"
                       variant="filled"
@@ -402,7 +420,7 @@ const MessageMcpTool: FC<Props> = ({ block }) => {
                       {t('common.cancel')}
                     </Button>
                   )}
-                  {isInvoking && toolResponse?.id ? (
+                  {isExecuting && toolResponse?.id ? (
                     <Button
                       size="small"
                       color="danger"
@@ -416,29 +434,31 @@ const MessageMcpTool: FC<Props> = ({ block }) => {
                       {t('chat.input.pause')}
                     </Button>
                   ) : (
-                    <StyledDropdownButton
-                      size="small"
-                      type="primary"
-                      icon={<ChevronDown size={14} />}
-                      onClick={() => {
-                        handleConfirmTool()
-                      }}
-                      menu={{
-                        items: [
-                          {
-                            key: 'autoApprove',
-                            label: t('settings.mcp.tools.autoApprove.label'),
-                            onClick: () => {
-                              handleAutoApprove()
+                    isWaitingConfirmation && (
+                      <StyledDropdownButton
+                        size="small"
+                        type="primary"
+                        icon={<ChevronDown size={14} />}
+                        onClick={() => {
+                          handleConfirmTool()
+                        }}
+                        menu={{
+                          items: [
+                            {
+                              key: 'autoApprove',
+                              label: t('settings.mcp.tools.autoApprove.label'),
+                              onClick: () => {
+                                handleAutoApprove()
+                              }
                             }
-                          }
-                        ]
-                      }}>
-                      <CirclePlay size={15} className="lucide-custom" />
-                      <CountdownText>
-                        {t('settings.mcp.tools.run', 'Run')} ({countdown}s)
-                      </CountdownText>
-                    </StyledDropdownButton>
+                          ]
+                        }}>
+                        <CirclePlay size={15} className="lucide-custom" />
+                        <CountdownText>
+                          {t('settings.mcp.tools.run', 'Run')} ({countdown}s)
+                        </CountdownText>
+                      </StyledDropdownButton>
+                    )
                   )}
                 </ActionButtonsGroup>
               </ActionsBar>
@@ -542,8 +562,7 @@ const ToolContentWrapper = styled.div`
     border: 1px solid var(--color-border);
   }
 
-  &.pending,
-  &.invoking {
+  &.pending {
     background-color: var(--color-background-soft);
     .ant-collapse {
       border: none;
@@ -663,6 +682,8 @@ const StatusIndicator = styled.span<{ status: string; hasError?: boolean }>`
         return 'var(--status-color-error)'
       case 'done':
         return props.hasError ? 'var(--status-color-error)' : 'var(--status-color-success)'
+      case 'error':
+        return 'var(--status-color-error)'
       default:
         return 'var(--color-text)'
     }
diff --git a/src/renderer/src/services/ApiService.ts b/src/renderer/src/services/ApiService.ts
index f48abed7c2..24225756a8 100644
--- a/src/renderer/src/services/ApiService.ts
+++ b/src/renderer/src/services/ApiService.ts
@@ -99,6 +99,8 @@ export async function fetchChatCompletion({
   const provider = AI.getActualProvider()
 
   const mcpTools: MCPTool[] = []
+  onChunkReceived({ type: ChunkType.LLM_RESPONSE_CREATED })
+
   if (isPromptToolUse(assistant) || isSupportedToolUse(assistant)) {
     mcpTools.push(...(await fetchMcpTools(assistant)))
   }
@@ -137,7 +139,6 @@ export async function fetchChatCompletion({
   }
 
   // --- Call AI Completions ---
-  onChunkReceived({ type: ChunkType.LLM_RESPONSE_CREATED })
   await AI.completions(modelId, aiSdkParams, {
     ...middlewareConfig,
     assistant,
diff --git a/src/renderer/src/services/WebSearchService.ts b/src/renderer/src/services/WebSearchService.ts
index c1ede59de3..fd6682fd39 100644
--- a/src/renderer/src/services/WebSearchService.ts
+++ b/src/renderer/src/services/WebSearchService.ts
@@ -178,12 +178,7 @@ class WebSearchService {
       formattedQuery = `today is ${dayjs().format('YYYY-MM-DD')} \r\n ${query}`
     }
 
-    // try {
     return await webSearchEngine.search(formattedQuery, websearch, httpOptions)
-    // } catch (error) {
-    //   console.error('Search failed:', error)
-    //   throw new Error(`Search failed: ${error instanceof Error ? error.message : 'Unknown error'}`)
-    // }
   }
 
   /**
diff --git a/src/renderer/src/services/messageStreaming/callbacks/baseCallbacks.ts b/src/renderer/src/services/messageStreaming/callbacks/baseCallbacks.ts
index a561ba937e..e7d7f005f7 100644
--- a/src/renderer/src/services/messageStreaming/callbacks/baseCallbacks.ts
+++ b/src/renderer/src/services/messageStreaming/callbacks/baseCallbacks.ts
@@ -69,16 +69,6 @@ export const createBaseCallbacks = (deps: BaseCallbacksDependencies) => {
       })
       await blockManager.handleBlockTransition(baseBlock as PlaceholderMessageBlock, MessageBlockType.UNKNOWN)
     },
-    // onBlockCreated: async () => {
-    //   if (blockManager.hasInitialPlaceholder) {
-    //     return
-    //   }
-    //   console.log('onBlockCreated')
-    //   const baseBlock = createBaseMessageBlock(assistantMsgId, MessageBlockType.UNKNOWN, {
-    //     status: MessageBlockStatus.PROCESSING
-    //   })
-    //   await blockManager.handleBlockTransition(baseBlock as PlaceholderMessageBlock, MessageBlockType.UNKNOWN)
-    // },
 
     onError: async (error: AISDKError) => {
       logger.debug('onError', error)
diff --git a/src/renderer/src/services/messageStreaming/callbacks/toolCallbacks.ts b/src/renderer/src/services/messageStreaming/callbacks/toolCallbacks.ts
index 1a5f3df28e..11a3641103 100644
--- a/src/renderer/src/services/messageStreaming/callbacks/toolCallbacks.ts
+++ b/src/renderer/src/services/messageStreaming/callbacks/toolCallbacks.ts
@@ -51,28 +51,6 @@ export const createToolCallbacks = (deps: ToolCallbacksDependencies) => {
       }
     },
 
-    onToolCallInProgress: (toolResponse: MCPToolResponse) => {
-      // 根据 toolResponse.id 查找对应的块ID
-      const targetBlockId = toolCallIdToBlockIdMap.get(toolResponse.id)
-
-      if (targetBlockId && toolResponse.status === 'invoking') {
-        const changes = {
-          status: MessageBlockStatus.PROCESSING,
-          metadata: { rawMcpToolResponse: toolResponse }
-        }
-        blockManager.smartBlockUpdate(targetBlockId, changes, MessageBlockType.TOOL)
-      } else if (!targetBlockId) {
-        logger.warn(
-          `[onToolCallInProgress] No block ID found for tool ID: ${toolResponse.id}. Available mappings:`,
-          Array.from(toolCallIdToBlockIdMap.entries())
-        )
-      } else {
-        logger.warn(
-          `[onToolCallInProgress] Received unhandled tool status: ${toolResponse.status} for ID: ${toolResponse.id}`
-        )
-      }
-    },
-
     onToolCallComplete: (toolResponse: MCPToolResponse) => {
       const existingBlockId = toolCallIdToBlockIdMap.get(toolResponse.id)
       toolCallIdToBlockIdMap.delete(toolResponse.id)
@@ -104,7 +82,6 @@ export const createToolCallbacks = (deps: ToolCallbacksDependencies) => {
             stack: null
           }
         }
-
         blockManager.smartBlockUpdate(existingBlockId, changes, MessageBlockType.TOOL, true)
 
         // Handle citation block creation for web search results
@@ -132,7 +109,6 @@ export const createToolCallbacks = (deps: ToolCallbacksDependencies) => {
           citationBlockId = citationBlock.id
           blockManager.handleBlockTransition(citationBlock, MessageBlockType.CITATION)
         }
-        // TODO: 处理 memory 引用
       } else {
         logger.warn(
           `[onToolCallComplete] Received unhandled tool status: ${toolResponse.status} for ID: ${toolResponse.id}`
diff --git a/src/renderer/src/store/thunk/__tests__/streamCallback.integration.test.ts b/src/renderer/src/store/thunk/__tests__/streamCallback.integration.test.ts
index 6c98e3cee4..96aff69f7b 100644
--- a/src/renderer/src/store/thunk/__tests__/streamCallback.integration.test.ts
+++ b/src/renderer/src/store/thunk/__tests__/streamCallback.integration.test.ts
@@ -458,18 +458,18 @@ describe('streamCallback Integration Tests', () => {
           }
         ]
       },
-      {
-        type: ChunkType.MCP_TOOL_IN_PROGRESS,
-        responses: [
-          {
-            id: 'tool-call-1',
-            tool: mockTool,
-            arguments: { testArg: 'value' },
-            status: 'invoking' as const,
-            response: ''
-          }
-        ]
-      },
+      // {
+      //   type: ChunkType.MCP_TOOL_PENDING,
+      //   responses: [
+      //     {
+      //       id: 'tool-call-1',
+      //       tool: mockTool,
+      //       arguments: { testArg: 'value' },
+      //       status: 'invoking' as const,
+      //       response: ''
+      //     }
+      //   ]
+      // },
       {
         type: ChunkType.MCP_TOOL_COMPLETE,
         responses: [
@@ -611,18 +611,18 @@ describe('streamCallback Integration Tests', () => {
           }
         ]
       },
-      {
-        type: ChunkType.MCP_TOOL_IN_PROGRESS,
-        responses: [
-          {
-            id: 'tool-call-1',
-            tool: mockCalculatorTool,
-            arguments: { operation: 'add', a: 1, b: 2 },
-            status: 'invoking' as const,
-            response: ''
-          }
-        ]
-      },
+      // {
+      //   type: ChunkType.MCP_TOOL_PENDING,
+      //   responses: [
+      //     {
+      //       id: 'tool-call-1',
+      //       tool: mockCalculatorTool,
+      //       arguments: { operation: 'add', a: 1, b: 2 },
+      //       status: 'invoking' as const,
+      //       response: ''
+      //     }
+      //   ]
+      // },
       {
         type: ChunkType.MCP_TOOL_COMPLETE,
         responses: [
diff --git a/src/renderer/src/utils/mcp-tools.ts b/src/renderer/src/utils/mcp-tools.ts
index c9ee86a76b..c9de9e1c50 100644
--- a/src/renderer/src/utils/mcp-tools.ts
+++ b/src/renderer/src/utils/mcp-tools.ts
@@ -826,9 +826,7 @@ export function mcpToolCallResponseToAwsBedrockMessage(
 }
 
 /**
- * 是否启用工具使用
- * 1. 如果模型支持函数调用，则启用工具使用
- * 2. 如果工具使用模式为 prompt，则启用工具使用
+ * 是否启用工具使用(function call)
  * @param assistant
  * @returns 是否启用工具使用
  */
diff --git a/yarn.lock b/yarn.lock
index 8c18baee73..0ebd717887 100644
--- a/yarn.lock
+++ b/yarn.lock
@@ -179,15 +179,15 @@ __metadata:
   languageName: node
   linkType: hard
 
-"@ai-sdk/google@npm:^2.0.7":
-  version: 2.0.7
-  resolution: "@ai-sdk/google@npm:2.0.7"
+"@ai-sdk/google@npm:^2.0.13":
+  version: 2.0.13
+  resolution: "@ai-sdk/google@npm:2.0.13"
   dependencies:
     "@ai-sdk/provider": "npm:2.0.0"
-    "@ai-sdk/provider-utils": "npm:3.0.4"
+    "@ai-sdk/provider-utils": "npm:3.0.8"
   peerDependencies:
     zod: ^3.25.76 || ^4
-  checksum: 10c0/bde4c95a2a167355cda18de9d5b273d562d2a724f650ca69016daa8df2766280487e143cf0cdd96f6654c255d587a680c6a937b280eb734ca2c35d6f9b9e943c
+  checksum: 10c0/a05210de11d7ab41d49bcd0330c37f4116441b149d8ccc9b6bc5eaa12ea42bae82364dc2cd09502734b15115071f07395525806ea4998930b285b1ce74102186
   languageName: node
   linkType: hard
 
@@ -227,15 +227,15 @@ __metadata:
   languageName: node
   linkType: hard
 
-"@ai-sdk/openai@npm:^2.0.19":
-  version: 2.0.19
-  resolution: "@ai-sdk/openai@npm:2.0.19"
+"@ai-sdk/openai@npm:^2.0.26":
+  version: 2.0.26
+  resolution: "@ai-sdk/openai@npm:2.0.26"
   dependencies:
     "@ai-sdk/provider": "npm:2.0.0"
-    "@ai-sdk/provider-utils": "npm:3.0.5"
+    "@ai-sdk/provider-utils": "npm:3.0.8"
   peerDependencies:
     zod: ^3.25.76 || ^4
-  checksum: 10c0/04db695669d783a810b80283e0cd48f6e7654667fd76ca2d35c7cffae6fdd68fb0473118e4e097ef1352f4432dd7c15c07f873d712b940c72495e5839b0ede98
+  checksum: 10c0/b8cb01c0c38525c38901f41f1693cd15589932a2aceddea14bed30f44719532a5e74615fb0e974eff1a0513048ac204c27456ff8829a9c811d1461cc635c9cc5
   languageName: node
   linkType: hard
 
@@ -267,20 +267,6 @@ __metadata:
   languageName: node
   linkType: hard
 
-"@ai-sdk/provider-utils@npm:3.0.5":
-  version: 3.0.5
-  resolution: "@ai-sdk/provider-utils@npm:3.0.5"
-  dependencies:
-    "@ai-sdk/provider": "npm:2.0.0"
-    "@standard-schema/spec": "npm:^1.0.0"
-    eventsource-parser: "npm:^3.0.3"
-    zod-to-json-schema: "npm:^3.24.1"
-  peerDependencies:
-    zod: ^3.25.76 || ^4
-  checksum: 10c0/4057810b320bda149a178dc1bfc9cdd592ca88b736c3c22bd0c1f8111c75ef69beec4a523f363e5d0d120348b876942fd66c0bb4965864da4c12c5cfddee15a3
-  languageName: node
-  linkType: hard
-
 "@ai-sdk/provider-utils@npm:3.0.7":
   version: 3.0.7
   resolution: "@ai-sdk/provider-utils@npm:3.0.7"
@@ -294,6 +280,19 @@ __metadata:
   languageName: node
   linkType: hard
 
+"@ai-sdk/provider-utils@npm:3.0.8":
+  version: 3.0.8
+  resolution: "@ai-sdk/provider-utils@npm:3.0.8"
+  dependencies:
+    "@ai-sdk/provider": "npm:2.0.0"
+    "@standard-schema/spec": "npm:^1.0.0"
+    eventsource-parser: "npm:^3.0.5"
+  peerDependencies:
+    zod: ^3.25.76 || ^4
+  checksum: 10c0/f466657c886cbb9f7ecbcd2dd1abc51a88af9d3f1cff030f7e97e70a4790a99f3338ad886e9c0dccf04dacdcc84522c7d57119b9a4e8e1d84f2dae9c893c397e
+  languageName: node
+  linkType: hard
+
 "@ai-sdk/provider@npm:2.0.0, @ai-sdk/provider@npm:^2.0.0":
   version: 2.0.0
   resolution: "@ai-sdk/provider@npm:2.0.0"
@@ -2249,8 +2248,8 @@ __metadata:
     "@ai-sdk/anthropic": "npm:^2.0.5"
     "@ai-sdk/azure": "npm:^2.0.16"
     "@ai-sdk/deepseek": "npm:^1.0.9"
-    "@ai-sdk/google": "npm:^2.0.7"
-    "@ai-sdk/openai": "npm:^2.0.19"
+    "@ai-sdk/google": "npm:^2.0.13"
+    "@ai-sdk/openai": "npm:^2.0.26"
     "@ai-sdk/openai-compatible": "npm:^1.0.9"
     "@ai-sdk/provider": "npm:^2.0.0"
     "@ai-sdk/provider-utils": "npm:^3.0.4"