Wenjuan Xing

https://www.wjx.cn/help/
Admin
Wenjuan Xing is an online survey and questionnaire platform that enables users to design,...
Tokens:336,573
Snippets:5,319
Trust Score:10
Update:5 months ago
Show doc for...
Tokens:
Raw
### Install CLI and Initialize

Source: https://www.wjx.cn/help/help.aspx?catid=136

Install the wjx-cli globally and initialize it with your API key. This is necessary for manual skill installation if npm is not available.

```bash
npm install -g wjx-cli
```

```bash
wjx init
```

--------------------------------

### Installation

Source: https://www.wjx.cn/help/help.aspx?catid=136

Install the wjx-api-sdk using npm or from source.

```APIDOC
## Installation

Install the wjx-api-sdk using npm:

```bash
npm install wjx-api-sdk
```

Or from source:

```bash
git clone https://github.com/wjxcom/wjx-ai-kit.git
cd wjx-ai-kit
npm install
npm run build --workspace=wjx-api-sdk
```

Requires Node.js >= 20.
```

--------------------------------

### Install and Configure wjx-cli

Source: https://www.wjx.cn/help/help.aspx?catid=136

Commands to install the wjx-cli globally, initialize it with your API key, and verify the connection.

```APIDOC
## Install CLI
npm install -g wjx-cli

## Configure API Key (Interactive, follow prompts)
wjx init

## Verify Connection
графииwjx doctor

## Install AI Skill Package to Current Directory
wjx skill install
```

--------------------------------

### Install Agent and Skills with CLI

Source: https://www.wjx.cn/help/help.aspx?catid=136

One-click installation of expert Agents (wjx-cli-expert) and associated skills using the 'npx wjx-cli skill install' command. This sets up the necessary directories and copies reference documentation.

```bash
npx wjx-cli skill install
```

--------------------------------

### Install wjx-cli

Source: https://www.wjx.cn/help/help.aspx?catid=136

Install the wjx-cli globally using npm. Ensure Node.js version is 20 or higher.

```bash
npm install -g wjx-cli
```

--------------------------------

### Install and Configure Wenjuanxing CLI with AI

Source: https://www.wjx.cn/help/help.aspx?catid=136

Instructions for an AI to install and configure the Wenjuanxing CLI (wjx-cli). It includes Node.js version checks, installation commands, API key retrieval for public and private deployments, initialization, and verification steps.

```shell
Please help me install and configure Wenjuanxing CLI (wjx-cli):

1. Check if Node.js is >= 20 (run `node --version`), if too low, guide me to upgrade at https://nodejs.org

2. Run `npm install -g wjx-cli` to install the CLI

3. Guide me to get the Wenjuanxing API Key:
   - Public cloud users: Ask me to visit https://www.wjx.cn/weixinlogin.aspx?redirecturl=%2Fnewwjx%2Fmanage%2Fuserinfo.aspx%3FshowApiKey%3D1, scan WeChat login, and copy the Key
   - Private deployment users: Replace www.wjx.cn with my domain (e.g., xxx.sojump.cn)
   Wait for me to send you the Key (possibly with a domain)

4. After getting the Key, run:
   `wjx init --api-key <My Key>`
   For private deployment users, add `--base-url https://<My Domain>`

5. Run `wjx doctor` to verify the connection, it should show API Key and network are ok

6. Finally, run `wjx survey list --table` to list my questionnaires. Seeing a real list means it's connected.
```

--------------------------------

### Manual Skill Package Installation

Source: https://www.wjx.cn/help/help.aspx?catid=136

Manually install the Skill package if npm is not available. This involves downloading, extracting, and then installing the CLI and configuring the API Key.

```bash
npm install -g wjx-cli  
wjx init
```

--------------------------------

### Install wjx-mcp-server

Source: https://www.wjx.cn/help/help.aspx?catid=136

Installs the wjx-mcp-server package globally, making it available for use with MCP-aware AI clients like Claude Desktop.

```bash
npm install -g wjx-mcp-server
```

--------------------------------

### MCP Server Docker Deployment

Source: https://www.wjx.cn/help/help.aspx?catid=136

Guide to building and running the MCP Server using Docker, including Docker Compose example.

```APIDOC
## Docker Deployment
### Build Image
Requires building from the monorepo root directory (as SDK is a workspace dependency):
  
```
git clone https://github.com/wjxcom/wjx-ai-kit.git
cd wjx-ai-kit
docker build -f wjx-mcp-server/Dockerfile -t wjx-mcp-server .
```

### Run
  
```
docker run -d \
  --name wjx-mcp \
  -p 3000:3000 \
  -e WJX_API_KEY=your_api_key \
  -e MCP_AUTH_TOKEN=your_secret \
  wjx-mcp-server
```

Includes health check (probes `/health` every 30 seconds), suitable for Kubernetes and Docker Compose.
### Docker Compose Example
  
```
services:
  wjx-mcp:
    build:
      context: .
      dockerfile: wjx-mcp-server/Dockerfile
    ports:
      - "3000:3000"
    environment:
      - WJX_API_KEY=${WJX_API_KEY}
      - MCP_AUTH_TOKEN=${MCP_AUTH_TOKEN}
    restart: unless-stopped
    healthcheck:
      test: ["CMD", "wget", "-q", "--spider", "http://localhost:3000/health"]
      interval: 30s
      timeout: 5s
      retries: 3
```
```

--------------------------------

### First API Call

Source: https://www.wjx.cn/help/help.aspx?catid=136

Example of making your first API calls to list surveys, get survey details, and create a survey using DSL text.

```APIDOC
## First API Call

```javascript
import { listSurveys, getSurvey, createSurveyByText } from "wjx-api-sdk";

// List surveys
const list = await listSurveys({ page_index: 1, page_size: 10 });
console.log(`Total ${list.data.total} surveys`);

// Get survey details
const survey = await getSurvey({ vid: 12345 });
console.log(survey.data.aname); // Survey title

// Create a survey using DSL text
const created = await createSurveyByText({
  text: `Customer Satisfaction Survey

1. Recommendation Intent [Scale Question]
0~10

2. Suggestions for Improvement [Text Question]`,
});
console.log(`New survey ID: ${created.data.vid}`);
```
```

--------------------------------

### Install wjx-api-sdk from Source

Source: https://www.wjx.cn/help/help.aspx?catid=136

Clone the repository and build the SDK from source if needed. Requires Node.js >= 20.

```bash
git clone https://github.com/wjxcom/wjx-ai-kit.git  
cd wjx-ai-kit  
npm install  
npm run build --workspace=wjx-api-sdk
```

--------------------------------

### Extract First Segment of Additional Information

Source: https://www.wjx.cn/help/help.aspx?helpid=776

Use this formula to get the first piece of information when segments are separated by '┋'. It finds the position of the first '┋' and extracts characters from the start up to that point.

```Formula
MID(A1,1,SEARCH("┋",A1,1)-1)
```

--------------------------------

### Quickly Integrate Claude Code with Questionnaire Star

Source: https://www.wjx.cn/help/help.aspx?catid=136

Automate the setup of Questionnaire Star MCP Server in Claude Code by sending a prompt. This includes version checks, API key retrieval, server addition, and skill installation.

```bash
请帮我接入问卷星 MCP Server：
1. 检查 node --version >= 20，过低引导我去 https://nodejs.org 升级
2. 让我去 https://www.wjx.cn/weixinlogin.aspx?redirecturl=%2Fnewwjx%2Fmanage%2Fuserinfo.aspx%3FshowApiKey%3D1 微信扫码取 API Key（私有化部署用户把 www.wjx.cn 换成自己的域名）
3. 拿到 Key 后执行：
   claude mcp add wjx --env WJX_API_KEY=<我的Key> -- npx -y wjx-mcp-server@latest
   私有化域名用户加 --env WJX_BASE_URL=<我的域名>
4. 执行 `claude mcp list` 确认 wjx 已注册
5. 提醒我**重启当前 Claude Code 会话**（关闭重开），然后调用 list_surveys 验证
6. 同时执行 `npx wjx-cli skill install` 安装专家 Agent + Skill 包
```

--------------------------------

### Manual Setup for Claude Code Integration

Source: https://www.wjx.cn/help/help.aspx?catid=136

Details the manual steps required for integrating with Questionnaire Star via Claude Code, including installing Claude Code, ensuring Node.js version compatibility, obtaining an API Key, and configuring the MCP server and agents.

```APIDOC
## Prerequisites

1.  **Install Claude Code** — `npm install -g @anthropic-ai/claude-code`, then run `claude` to authenticate.
2.  **Node.js >= 20** — Run `node --version` to confirm. If not installed or version is below 20, installation is recommended via:
    *   **Using nvm (Recommended)**: `nvm install 20 && nvm use 20` (nvm installation instructions; Windows users can use nvm-windows).
    *   **Official Download**: Go to nodejs.org to download the LTS version installer.
3.  **Obtain Questionnaire Star API Key** — Log in via QR code on the API Key acquisition page; the API Key will be displayed directly on the page.

## Step 1: Integrate MCP Server

### Method 1: Command Line Addition (Recommended)

```bash
# Basic configuration
claude mcp add wjx --env WJX_API_KEY=YOUR_API_KEY -- npx -y wjx-mcp-server@latest

# Enterprise users (for contact list functionality)
claude mcp add wjx --env WJX_API_KEY=YOUR_API_KEY --env WJX_CORP_ID=YOUR_CORP_ID -- npx -y wjx-mcp-server@latest
```

### Method 2: Manually Edit Configuration
Create a `.mcp.json` file in your project directory:

```json
{
  "mcpServers": {
    "wjx": {
      "command": "npx",
      "args": ["-y", "wjx-mcp-server@latest"],
      "env": {
        "WJX_API_KEY": "Replace with your API Key"
      }
    }
  }
}
```

### HTTP Remote Mode
When a remote MCP service is deployed by your team:

```json
{
  "mcpServers": {
    "wjx": {
      "url": "http://your-server:3000/mcp",
      "headers": {
        "Authorization": "Bearer your-token"
      }
    }
  }
}
```

### Verify MCP Connection

```bash
claude mcp list
# Should show the wjx server and its tool list
```

## Step 2: Deploy Agent (Recommended)
Agents are a core advantage of wjx-ai-kit over other solutions. Two expert agents include complete questionnaire operation workflows:
*   **wjx-mcp-expert** — MCP tool expert, skilled in operating Questionnaire Star via MCP protocol.
*   **wjx-cli-expert** — CLI command expert, skilled in batch operations via command line.

### Method 1: One-Click Installation using CLI (Recommended)

```bash
npx wjx-cli skill install
```

This command automatically performs the following:
*   Creates the `.claude/agents/` directory and deploys the `wjx-cli-expert` Agent (CLI Command Expert).
*   Copies the `skills/wjx-cli-use/` reference documentation (DSL syntax, question type encoding, CLI command parameters, etc.).

> **Note**: For the MCP tool expert Agent (`wjx-mcp-expert`), use Method 2 to install from the Skill market, or Method 3 for manual deployment.
### Method 2: Install from Skill Market
`wjx-ai-kit` Skills are available on Vercel Agent Skills and ClawHub markets:

```bash
# Install via Vercel Agent Skills
npx skills add wjxcom/wjx-ai-kit
```

### Method 3: Manual Deployment
Download the following files from the GitHub repository to your project directory:

```
# Files to download:
wjx-agents/wjx-mcp-expert/wjx-mcp-expert.md  →  .claude/agents/wjx-mcp-expert.md
wjx-agents/wjx-cli-expert/wjx-cli-expert.md  →  .claude/agents/wjx-cli-expert.md

# Download the entire directory:
wjx-skills/wjx-cli-use/                      →  skills/wjx-cli-use/
wjx-skills/wjx-mcp-use/                      →  skills/wjx-mcp-use/
```

### Method 4: Manual Installation of Skill Zip
If npm is not available, download the Skill package and install manually:
1.  Download `wjx-cli-use-skill-latest.zip`.
2.  Unzip and extract the `wjx-cli-use` directory into the `skills/` directory of your project.
3.  Install CLI and configure API Key:
    ```bash
    npm install -g wjx-cli
    wjx init
    ```

See the Skill package getting started guide for details.
```

--------------------------------

### Initialize wjx-cli (Interactive)

Source: https://www.wjx.cn/help/help.aspx?catid=136

Configure wjx-cli interactively by running the init command. This is recommended for new users.

```bash
wjx init
```

--------------------------------

### AI Prompt for MCP Server Integration

Source: https://www.wjx.cn/help/help.aspx?catid=136

This prompt guides an AI assistant to help users integrate with Questionnaire Star's MCP Server. It covers checking Node.js version, obtaining an API Key, configuring the MCP server for different AI clients (Claude Code, others), and verifying the setup.

```text
请帮我接入问卷星 MCP Server：  
  
1. 检查 Node.js 是否 ≥ 20，版本过低引导我去 https://nodejs.org 升级  
2. 让我去 https://www.wjx.cn/weixinlogin.aspx?redirecturl=%2Fnewwjx%2Fmanage%2Fuserinfo.aspx%3FshowApiKey%3D1 微信扫码取 API Key（私有化部署用户把 www.wjx.cn 换成自己的域名），等我把 Key 发给你  
3. 根据我的 AI 工具自动选配置方式：  
   - Claude Code：执行 `claude mcp add wjx --env WJX_API_KEY=<我的Key> -- npx -y wjx-mcp-server@latest`  
   - 其他 MCP 客户端：写入它们的 MCP 配置文件  
     ```json  
     {"mcpServers":{"wjx":{"command":"npx","args":["-y","wjx-mcp-server@latest"],"env":{"WJX_API_KEY":"<我的Key>"}}}}  
     ```  
4. 配置完提醒我**完全重启** AI 工具  
5. 重启后调用 `list_surveys` 验证
```

--------------------------------

### First SDK API Calls: List, Get, Create Survey

Source: https://www.wjx.cn/help/help.aspx?catid=136

Demonstrates basic usage of the SDK by listing surveys, retrieving a specific survey's details, and creating a new survey using DSL text.

```javascript
import { listSurveys, getSurvey, createSurveyByText } from "wjx-api-sdk";  

// 列出问卷  
const list = await listSurveys({ page_index: 1, page_size: 10 });  
console.log(`共 ${list.data.total} 份问卷`);  

// 获取问卷详情  
const survey = await getSurvey({ vid: 12345 });  
console.log(survey.data.aname); // 问卷标题  

// 用 DSL 文本创建问卷  
const created = await createSurveyByText({
  text: `客户满意度调查  
  
1. 推荐意愿 [量表题]  
0~10  
  
2. 改进建议 [填空题]`,  
});  
console.log(`新问卷 ID: ${created.data.vid}`);
```

--------------------------------

### Verify wjx-cli Installation

Source: https://www.wjx.cn/help/help.aspx?catid=136

Check the installed version of wjx-cli to confirm successful installation.

```bash
wjx --version
```

--------------------------------

### Manually Install Skill Zip Package

Source: https://www.wjx.cn/help/help.aspx?catid=136

Install the Skill package manually by downloading the zip file, extracting the 'wjx-cli-use' directory into the 'skills/' folder, and then initializing the CLI with your API Key.

```bash
npm install -g wjx-cli
графии init
```
Tokens:
Raw