## 更新内容
* feat(cli/sdk)：在非交互模式和 SDK API 中暴露 /context 使用数据 by @wenshao in https://github.com/QwenLM/qwen-code/pull/2916
* feat(cli)：添加启动性能分析器 by @doudouOUC in https://github.com/QwenLM/qwen-code/pull/3232
* feat(core)：实现用于上下文共享的 fork 子代理 by @wenshao in https://github.com/QwenLM/qwen-code/pull/2936
* feat(vscode-ide-companion)：添加 /account 命令以显示账户信息 by @qqqys in https://github.com/QwenLM/qwen-code/pull/2984
* fix(core)：检测来自流式 SSE 帧的速率限制错误 by @wenshao in https://github.com/QwenLM/qwen-code/pull/3246
* fix(vscode)：限制会话标签页标题长度以防止标签栏溢出 by @dreamWB in https://github.com/QwenLM/qwen-code/pull/3249
* fix(core)：遵循 modelProviders 配置中的自定义 Gemini baseUrl by @doudouOUC in https://github.com/QwenLM/qwen-code/pull/3212
* fix(core)：在 GeminiChat 流验证中允许仅含思考过程的响应 by @wenshao in https://github.com/QwenLM/qwen-code/pull/3251
* fix(cli)：使 /bug 命令在无超链接支持的终端中更易打开 by @reidliu41 in https://github.com/QwenLM/qwen-code/pull/3257
* feat(acp)：基于 LLM 的自定义提示词消息重写中间件 by @zhangxy-zju in https://github.com/QwenLM/qwen-code/pull/3191
* fix(cli)：在 BaseTextInput 中忽略字面量 Tab 键输入 by @Shenean in https://github.com/QwenLM/qwen-code/pull/3270
* feat(auth)：停用 Qwen OAuth 免费套餐（2026-04-15 截止） by @tanzhenxin in https://github.com/QwenLM/qwen-code/pull/3291
* fix(channels/dingtalk)：在 allowedUsers 匹配中优先使用 senderStaffId 而非 senderId by @yiliang114 in https://github.com/QwenLM/qwen-code/pull/3294
* chore(release)：将版本号升级至 0.14.5 by @tanzhenxin in https://github.com/QwenLM/qwen-code/pull/3298
* fix(cli)：在 ModelDialog 中屏蔽已停用的 qwen-oauth 模型选项 by @tanzhenxin in https://github.com/QwenLM/qwen-code/pull/3299
## 新贡献者
* @dreamWB 在 https://github.com/QwenLM/qwen-code/pull/3249 中完成首次贡献
* @reidliu41 在 https://github.com/QwenLM/qwen-code/pull/3257 中完成首次贡献
* @Shenean 在 https://github.com/QwenLM/qwen-code/pull/3270 中完成首次贡献
**完整更新日志**: https://github.com/QwenLM/qwen-code/compare/v0.14.4...v0.14.5
## What's Changed
* feat(cli/sdk): expose /context usage data in non-interactive mode and SDK API by @wenshao in https://github.com/QwenLM/qwen-code/pull/2916
* feat(cli): add startup performance profiler by @doudouOUC in https://github.com/QwenLM/qwen-code/pull/3232
* feat(core): implement fork subagent for context sharing by @wenshao in https://github.com/QwenLM/qwen-code/pull/2936
* feat(vscode-ide-companion): add /account for account display by @qqqys in https://github.com/QwenLM/qwen-code/pull/2984
* fix(core): detect rate-limit errors from streamed SSE frames by @wenshao in https://github.com/QwenLM/qwen-code/pull/3246
* fix(vscode): limit session tab title length to prevent tab bar overflow by @dreamWB in https://github.com/QwenLM/qwen-code/pull/3249
* fix(core): respect custom Gemini baseUrl from modelProviders by @doudouOUC in https://github.com/QwenLM/qwen-code/pull/3212
* fix(core): allow thought-only responses in GeminiChat stream validation by @wenshao in https://github.com/QwenLM/qwen-code/pull/3251
* fix(cli): make /bug easier to open in terminals without hyperlink support by @reidliu41 in https://github.com/QwenLM/qwen-code/pull/3257
* feat(acp): LLM-based message rewrite middleware with custom prompts by @zhangxy-zju in https://github.com/QwenLM/qwen-code/pull/3191
* fix(cli): ignore literal Tab input in BaseTextInput by @Shenean in https://github.com/QwenLM/qwen-code/pull/3270
* feat(auth): discontinue Qwen OAuth free tier (2026-04-15 cutoff) by @tanzhenxin in https://github.com/QwenLM/qwen-code/pull/3291
* fix(channels/dingtalk): prioritize senderStaffId over senderId for allowedUsers matching by @yiliang114 in https://github.com/QwenLM/qwen-code/pull/3294
* chore(release): bump version to 0.14.5 by @tanzhenxin in https://github.com/QwenLM/qwen-code/pull/3298
* fix(cli): block discontinued qwen-oauth model selection in ModelDialog by @tanzhenxin in https://github.com/QwenLM/qwen-code/pull/3299
## New Contributors
* @dreamWB made their first contribution in https://github.com/QwenLM/qwen-code/pull/3249
* @reidliu41 made their first contribution in https://github.com/QwenLM/qwen-code/pull/3257
* @Shenean made their first contribution in https://github.com/QwenLM/qwen-code/pull/3270
**Full Changelog**: https://github.com/QwenLM/qwen-code/compare/v0.14.4...v0.14.5