[add] Prototype Generator model & component

[Copilot]

实现原型生成器工具栏组件

Implementation of a "Prototype Generator Toolbar" component that displays after each AI evaluation message.

Changes Made:

1. Converted to class component - Rewrote using ObservedComponent and MobX instead of React hooks
2. Removed custom method - Using base class's updateOne() instead of custom createVersion()
3. Removed unified exports - Components imported directly from specific directories
4. Removed long documentation - Deleted .md file, code is self-documenting
5. Split render methods - Separated renderContent into renderPending(), renderGenerating(), renderCompleted(), and renderFailed() for better code organization
6. Improved polling mechanism - Replaced setInterval with for...of loop + await sleep() pattern for cleaner async flow control

Technical Details:

Component Structure:

• Uses ObservedComponent from mobx-react-helper
• Uses @observable decorator for reactive state
• Implements lifecycle methods (componentDidMount, componentDidUpdate, componentWillUnmount)
• Separate render methods for each state (Pending, Generating, Completed, Failed)

Polling Mechanism:

• Uses for...of loop with await sleep(3) for cleaner async control
• Observable isPolling flag for proper state management
• Automatic cleanup on component unmount
• Maximum 100 iterations (~5 minutes) before stopping

API Integration:

• Uses base updateOne() for creating prototype versions: updateOne({ messageId })
• Custom method getVersionByMessageId() for fetching by message

4 States:

1. Pending - Generate button
2. Generating - Loading spinner with auto-polling
3. Completed - Preview and AI log links
4. Failed - Error message with AI log link

Original prompt

实现一个“原型生成器工具栏”组件，然后插入到每条 AI 评估消息后，并具备以下几种状态：

1. 待生成，可点击按钮发起原型版本生成请求
2. 生成中，GET 轮询原型版本信息，不断更新生成状态
3. 已生成，可点击查看预览链接、AI 日志
4. 未生成，显示错误消息，也可点击查看 AI 日志

其中后端接口使用 https://github.com/idea2app/data-server/pull/24 ，数据类型定义从最新版 @idea2app/data-server 导入。

Human modification

1. [migrate] replace SMS Code with Email OTP for sign-in
2. [refactor] simplify Status Check polling of Prototype Generation
3. [optimize] move Prototype Generator to Evaluation Scope lines
4. [optimize] update Upstream packages
5. [fix] many detail bugs

---

💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.

[TechQuery]

@copilot

[TechQuery]

@copilot