LMCanvas

UI generated by LLM, Interact with LLM

基于大模型的UI动态生成，UI与大模型交互

This is a prototype, there are many imperfections.

这只是一个原型，还有很多不完善的地方

与众不同之处 / Unique Features

相对于Code Preview(Gemini的Canvas、Claude的Artifacts)，它不再是单次生成单次预览，而是继承式的补充；它不是用户分别与模型和预览交互，而是在用户与界面交互时界面与模型交互

Compared with Code Preview(Gemini's Cnavas/Claude's Artifacts), it is not a one-time generation and preview, but an incremental supplement; it is not that the user interacts with the model and preview separately, but that the interface interacts with the model when the user interacts with the interface.

功能特性 / Features

• 由LLM实时生成代码(兼容OpenAI API)
• Codes generated by LLM instantly(compatible with OpenAI API)
• 通过asak混用模型，避免触发速率限制
• Mix models using asak, avoiding rate limiting
• 基于Fluent UI React Components构建的美观控制台界面
• Fluent UI React Components-based beautiful console interface
• 代码高亮和Markdown支持
• Code highlighting and Markdown support

安装 / Installation

直接使用 / Direct Use

前往Releases页面下载Windows版本安装包

Go to the Releases page to download the Windows installation package.

通过源代码构建 / Build from Source

下载安装node.js：官网

1. Download and install node.js: official website

2. 克隆仓库 / Clone the repository

   git clone git@github.com:for-the-zero/LMCanvas.git

3. 安装依赖 / Install dependencies

   npm install --verbose

4. 构建 / Build

   npm run build

5. 启动 / Start

   构建产物在builds文件夹内

   The build product is in the builds folder.

AI能做到的 / What AI can do

• 生成页面
• Generate a pages
• 通过send2ai函数实现界面与LLM交互
• Implement interface interaction with LLM through the send2ai function
• 使用node的fs path模块和electron的clipboard fetch shell模块
• Use the fs path module and clipboard fetch shell modules of node and electron respectively

系统提示词 / System prompts

由Gemini 2.5 Pro根据Prompt 101

Generated by Gemini 2.5 Pro based on Prompt 101

相关链接 / Links

• GitHub仓库 / GitHub Repository
• 免费LLM API集合 / Free LLM Collection
• asak
• Fluent UI React Components

图标 / Icon

Generated by AppIcon Forge

Author: Streamline

Licence: CC BY 4.0