Miniprogram Development

WeChat Mini Program Development Rules

Project Structure

1. CloudBase Integration:

   • If user needs to develop mini program, you will use various WeChat CloudBase capabilities to develop the project
   • Mini program base library should use latest version

2. Directory Organization:

   • Mini program projects follow WeChat CloudBase best practices
   • Mini program code is generally in miniprogram directory
   • If developing cloud functions, they can be stored in cloudfunctions directory
   • Mini program's project.config.json needs to specify miniprogramRoot and other configurations

3. Page Generation:

   • When generating mini program pages, must include page configuration files such as index.json
   • Must comply with specifications to avoid compilation errors

Development Tools

WeChat Developer Tools Opening Project Workflow:

• When detecting current project is a mini program project, suggest user to use WeChat Developer Tools for preview, debugging, and publishing
• Before opening, confirm project.config.json has appid field configured. If not configured, must ask user to provide it
• Use WeChat Developer built-in CLI command to open project (pointing to directory containing project.config.json):
  • Windows: "C:\Program Files (x86)\Tencent\微信web开发者工具\cli.bat" open --project "项目根目录路径"
  • macOS: /Applications/wechatwebdevtools.app/Contents/MacOS/cli open --project "/path/to/project/root"
• Project root directory path is the directory containing project.config.json file

CloudBase Integration

1. Environment Configuration:

   • When using wx.cloud in mini program, need to specify environment ID
   • Environment ID can be queried via envQuery tool

2. Resource Management:

   • When generating mini program code, if material images are needed, such as tabbar's iconPath and other places, prefer Icons8 (see section 4 above for details)
   • Use downloadRemoteFile tool to download resources
   • When generating mini program code, if using iconPath and similar, must simultaneously help user download icons to avoid build errors

Mini Program Authentication Characteristics

Important: Mini programs with CloudBase are naturally login-free. It is strictly forbidden to generate login pages or login flows!

1. Login-Free Feature: Mini program CloudBase does not require user login, can get user identity in cloud functions via wx-server-sdk

2. User Identity Retrieval: In cloud functions, get user's unique identifier via cloud.getWXContext().OPENID

3. User Data Management: Manage user data in cloud functions based on openid, no login flow needed

// Example of getting user identity in cloud function
exports.main = async (event, context) => {
  const wxContext = cloud.getWXContext();
  const openid = wxContext.OPENID;
  
  return { openid: openid };
};

AI Model Invocation

Mini programs with base library version 3.7.1+ already support direct AI model invocation

// Create model instance, here we use DeepSeek AI model
const model = wx.cloud.extend.AI.createModel("deepseek");

// First set AI's system prompt, here using seven-character quatrain generation as example
const p =
  "请严格按照七言绝句或七言律诗的格律要求创作，平仄需符合规则，押韵要和谐自然，韵脚字需在同一韵部。创作内容围绕用户给定的主题，七言绝句共四句，每句七个字；七言律诗共八句，每句七个字，颔联和颈联需对仗工整。同时，要融入生动的意象、丰富的情感与优美的意境，展现出古诗词的韵味与美感。";

// User's natural language input, e.g., '帮我写一首赞美玉龙雪山的诗'
const userInput = "帮我写一首赞美玉龙雪山的诗";

// Pass system prompt and user input to AI model
const res = await model.streamText({
  data: {
    model: "deepseek-v3", // Specify specific model
    messages: [
      { role: "system", content: p },
      { role: "user", content: userInput },
    ],
  },
});

// Receive AI model's response
// Since AI model's return result is streaming, we need to loop to receive complete response text
for await (let str of res.textStream) {
  console.log(str);
}

WeChat Step Count Retrieval

WeChat step count retrieval must use CloudID method (base library 2.7.0+):

• Frontend: wx.getWeRunData() to get cloudID, use wx.cloud.CloudID(cloudID) to pass to cloud function
• Cloud Function: Directly use weRunData.data to get decrypted step count data, check weRunData.errCode to handle errors
• Forbidden: Do not use session_key manual decryption method, CloudID is more secure and simple
• Required: Must implement fallback mechanism (simulated data) to handle cloudID retrieval failure cases

Cloud Function Deployment and Permission Notes

• After AI automatically deploys cloud functions, special permissions like cloud invocation may be missing
• Recommend users to right-click cloud function in WeChat Developer Tools, select "Install Dependencies in Cloud"
• For functions requiring cloud invocation permissions (such as WeChat step count decryption), recommend manually deploying once via Developer Tools to get complete permissions
• If encountering permission issues, prompt user to check cloud function's service authorization and API permission configuration

Development Workflow Guidance

• After completing mini program project development, proactively suggest user to use WeChat Developer Tools for preview, debugging, and publishing
• If user agrees, use CLI command to open WeChat Developer Tools, pointing to project root directory containing project.config.json
• Remind user to perform real device preview, debugging, and publishing operations in WeChat Developer Tools