codeg v0.20.0

Release version 0.20.0

• feat(acp): Meet Grok Build — the 11th built-in agent. xAI's Grok Build joins codeg as a first-class coding agent: sign in with your xAI API key, choose its model and reasoning effort, delegate sub-agents, and manage its MCP servers — with live and replayed sessions rendered exactly like every other agent.
• feat(science): A new Scientific Research skills mode. Thirteen curated, open-source research skills (from K-Dense-AI's scientific-agent-skills) now ship alongside the bundled expert and office skills. Install them to the shared library, enable them per agent in Settings, and start from the new "Scientific Research" tab on the new-conversation screen — localized in all 10 languages.
• feat(chat): The composer and your sent messages are now plain text. What you type, and the messages you've already sent, display as clean plain text with file and session references shown as compact badges — so your input reads exactly as written instead of being reflowed as Markdown. (Assistant replies are unchanged and still render Markdown.)
• fix(chat): Background tasks now settle correctly. A background task that's been collected or stopped is now marked as finished, so its indicator no longer lingers as if it were still running.
• fix(chat): The model picker stays open while you use its scrollbar. Dragging the scrollbar inside the model list no longer dismisses the picker.

---

发布版本 0.20.0

• 功能(智能体)：认识第 11 个内置智能体——Grok Build。 xAI 的 Grok Build 作为一等编码智能体加入 codeg：用 xAI API Key 登录，选择模型与推理强度，委托子智能体，管理它的 MCP 服务器——实时与回放会话的渲染都与其它智能体完全一致。
• 功能(科研)：全新「科学研究」技能模式。 13 个精选的开源科研技能（来自 K-Dense-AI 的 scientific-agent-skills）现已与内置的专家技能、办公技能一同发布。可安装到共享技能库、在设置中按智能体启用，并从新会话页面新增的「科学研究」标签开始——已本地化为全部 10 种语言。
• 功能(聊天)：输入框和你发送的消息现在为纯文本。 你正在输入的内容以及已发送的消息，都以清爽的纯文本显示，文件与会话引用呈现为紧凑徽章——所见即所写，不再被重排为 Markdown。（智能体的回复不受影响，仍以 Markdown 渲染。）
• 修复(聊天)：后台任务现在能正确结算。 被收集或停止的后台任务现在会被标记为已完成，其指示条不再一直停留在「运行中」。
• 修复(聊天)：拖动滚动条时模型选择器保持打开。 在模型列表内拖动滚动条不再关闭选择器。