<!-- qwen-release-notes:v1 -->
## 亮点
- 为 macOS 用户新增可选的一次性更新桥接功能，以便从 Electron 桌面应用迁移至全新的 Tauri 壳。 ([#8392](https://github.com/QwenLM/qwen-code/pull/8392))
- 引入针对工具调用的详细执行结果追踪，以区分调用成功、失败和取消。 ([#8180](https://github.com/QwenLM/qwen-code/pull/8180))
- 除非工具明确标记为幂等且工作区可信，否则在连接断开后防止不安全地重放 MCP 工具调用。 ([#8387](https://github.com/QwenLM/qwen-code/pull/8387))
- 在证据目录耗尽后，通过将目标（Goal）切换至需要用户干预的 usage_limited 状态，停止无限重试 Goal 模式。 ([#8430](https://github.com/QwenLM/qwen-code/pull/8430))
- 在 Web Shell 中引入以发现项为核心的结构化审查结果视图，展示严重程度、置信度及建议修复方案。 ([#8402](https://github.com/QwenLM/qwen-code/pull/8402))
- 在增强型 Markdown 表格中关闭单元格值对话框时，防止页面意外滚动，同时保留键盘焦点。 ([#8407](https://github.com/QwenLM/qwen-code/pull/8407))
## 破坏性变更
暂无已知的破坏性变更。
## 完整变更列表
### 新功能
- 为 macOS 用户新增可选的一次性更新桥接功能，以便从 Electron 桌面应用迁移至全新的 Tauri 壳。 ([#8392](https://github.com/QwenLM/qwen-code/pull/8392)) by @yiliang114
- 引入针对工具调用的详细执行结果追踪，以区分调用成功、失败和取消。 ([#8180](https://github.com/QwenLM/qwen-code/pull/8180)) by @doudouOUC
- 在 Web Shell 中引入以发现项为核心的结构化审查结果视图，展示严重程度、置信度及建议修复方案，并附带持久化报告。 ([#8402](https://github.com/QwenLM/qwen-code/pull/8402)) by @wenshao
- 通过将现有努力标志映射到提供商特定的 reasoning_effort 参数，启用对 Qwen 3.8 推理努力级别的支持。 ([#8472](https://github.com/QwenLM/qwen-code/pull/8472)) by @DragonnZhang
### 缺陷修复
- 在增强型 Markdown 表格中关闭单元格值对话框时，防止页面意外滚动，同时保留键盘焦点。 ([#8407](https://github.com/QwenLM/qwen-code/pull/8407)) by @carffuca
- 明确 write_file 指导说明，当目标文件的存在状态或内容未知时，显式要求先读取再写入。 ([#8428](https://github.com/QwenLM/qwen-code/pull/8428)) by @doudouOUC
- 通过从 Tauri 运行时使用的正确 LocalAppData 路径读取日志，修复 Windows 桌面版发布的冒烟测试。 ([#8381](https://github.com/QwenLM/qwen-code/pull/8381)) by @yiliang114
- 除非工具明确标记为幂等且工作区可信，否则在连接断开后防止不安全地重放 MCP 工具调用。 ([#8387](https://github.com/QwenLM/qwen-code/pull/8387)) by @doudouOUC
- 在证据目录耗尽后，通过将目标（Goal）切换至需要用户干预的 usage_limited 状态，停止无限重试 Goal 模式。 ([#8430](https://github.com/QwenLM/qwen-code/pull/8430)) by @qqqys
- 防止审查系统对已成功交付的发布任务不必要地重新运行代理，从而节省时间和 Token。 ([#8466](https://github.com/QwenLM/qwen-code/pull/8466)) by @wenshao
- 修复 MCP 重连的失败单元测试，确保其能正确验证在可信服务器配置下的安全重放策略。 ([#8478](https://github.com/QwenLM/qwen-code/pull/8478)) by @wenshao
- 更新审查工作流测试，使其正确断言外部化的超时变量，解决质量检查作业中的失败问题。 ([#8486](https://github.com/QwenLM/qwen-code/pull/8486)) by @wenshao
### 内部改进
- 移除损坏的旧版定时 PR 分类工作流，该工作流此前因错误解析问题引用而无法同步标签。 ([#8434](https://github.com/QwenLM/qwen-code/pull/8434)) by @wenshao
- 通过在新提交时取消过期运行并将可信 Linux 作业路由至更快的 ECS 运行器，缩短 SDK Java 测试的排队时间。 ([#8441](https://github.com/QwenLM/qwen-code/pull/8441)) by @yiliang114
**完整更新日志**: https://github.com/QwenLM/qwen-code/compare/v0.21.4...v0.21.5
<!-- qwen-release-notes:v1 -->
## Highlights
- Adds an opt-in one-time update bridge for macOS users to migrate from the Electron desktop app to the new Tauri shell. ([#8392](https://github.com/QwenLM/qwen-code/pull/8392))
- Introduces detailed execution-specific outcome tracking for tool calls to distinguish between invocation success, failure, and cancellation. ([#8180](https://github.com/QwenLM/qwen-code/pull/8180))
- Prevents unsafe replay of MCP tool calls after connection loss unless the tool is explicitly marked idempotent and the workspace is trusted. ([#8387](https://github.com/QwenLM/qwen-code/pull/8387))
- Stops infinite Goal mode retries after evidence catalog exhaustion by transitioning the Goal to a usage_limited state requiring user intervention. ([#8430](https://github.com/QwenLM/qwen-code/pull/8430))
- Introduces a structured, finding-centric view in Web Shell for review results, displaying severity, confidence, and suggested fixes. ([#8402](https://github.com/QwenLM/qwen-code/pull/8402))
- Prevents unexpected page scrolling when closing cell-value dialogs in enhanced Markdown tables while preserving keyboard focus. ([#8407](https://github.com/QwenLM/qwen-code/pull/8407))
## Breaking Changes
No known breaking changes.
## Complete Change List
### Features
- Adds an opt-in one-time update bridge for macOS users to migrate from the Electron desktop app to the new Tauri shell. ([#8392](https://github.com/QwenLM/qwen-code/pull/8392)) by @yiliang114
- Introduces detailed execution-specific outcome tracking for tool calls to distinguish between invocation success, failure, and cancellation. ([#8180](https://github.com/QwenLM/qwen-code/pull/8180)) by @doudouOUC
- Introduces a structured, finding-centric view in Web Shell for review results, displaying severity, confidence, and suggested fixes alongside durable reports. ([#8402](https://github.com/QwenLM/qwen-code/pull/8402)) by @wenshao
- Enables support for Qwen 3.8 reasoning effort levels by mapping existing effort flags to the provider-specific reasoning_effort parameter. ([#8472](https://github.com/QwenLM/qwen-code/pull/8472)) by @DragonnZhang
### Bug Fixes
- Prevents unexpected page scrolling when closing cell-value dialogs in enhanced Markdown tables while preserving keyboard focus. ([#8407](https://github.com/QwenLM/qwen-code/pull/8407)) by @carffuca
- Clarifies write_file guidance to explicitly require reading a target file before writing when its existence or contents are unknown. ([#8428](https://github.com/QwenLM/qwen-code/pull/8428)) by @doudouOUC
- Fixes the Windows desktop release smoke test by reading logs from the correct LocalAppData path used by the Tauri runtime. ([#8381](https://github.com/QwenLM/qwen-code/pull/8381)) by @yiliang114
- Prevents unsafe replay of MCP tool calls after connection loss unless the tool is explicitly marked idempotent and the workspace is trusted. ([#8387](https://github.com/QwenLM/qwen-code/pull/8387)) by @doudouOUC
- Stops infinite Goal mode retries after evidence catalog exhaustion by transitioning the Goal to a usage_limited state requiring user intervention. ([#8430](https://github.com/QwenLM/qwen-code/pull/8430)) by @qqqys
- Prevents the review system from unnecessarily re-running agents for launches that were already successfully delivered, saving time and tokens. ([#8466](https://github.com/QwenLM/qwen-code/pull/8466)) by @wenshao
- Fixes a failing unit test for MCP reconnection to ensure it correctly validates safe replay policies with trusted server configurations. ([#8478](https://github.com/QwenLM/qwen-code/pull/8478)) by @wenshao
- Updates review workflow tests to correctly assert against externalized timeout variables, resolving failures in the Quality Checks job. ([#8486](https://github.com/QwenLM/qwen-code/pull/8486)) by @wenshao
### Internal Changes
- Removes the broken legacy scheduled PR triage workflow that previously failed to sync labels due to incorrect issue reference parsing. ([#8434](https://github.com/QwenLM/qwen-code/pull/8434)) by @wenshao
- Reduces queue times for SDK Java tests by canceling stale runs on new commits and routing trusted Linux jobs to faster ECS runners. ([#8441](https://github.com/QwenLM/qwen-code/pull/8441)) by @yiliang114
**Full Changelog**: https://github.com/QwenLM/qwen-code/compare/v0.21.4...v0.21.5