← 返回 Skills 市场

超星作业自动批改

Name: 超星作业自动批改
Author: chenghaozhangswu

作者 chenghaozhangswu · GitHub ↗ · v1.0.0 · MIT-0

cross-platform ⚠ suspicious

226

总下载

当前安装

版本数

在 OpenClaw 中安装

/install chaoxing-auto-grade

功能描述

自动登录超星学习通，智能识别未批改主观题并调用通义千问AI完成评分及成绩自动提交。

使用说明 (SKILL.md)

Chaoxing Auto Grade

超星（学习通）作业自动批改技能。使用 Playwright 自动化 + 通义千问 AI 进行主观题智能评分。

功能

🔐 自动登录学习通
📚 根据课程名选择课程
📝 自动找到未批改的作业
🤖 AI 智能评分（通义千问 API）
✅ 自动提交成绩
📄 支持翻页处理多个作业

配置

修改 config.json：

{
  "username": "你的学习通账号",
  "password": "你的学习通密码",
  "courseName": "课程名称（留空选第一个）",
  "apiKey": "通义千问 API Key",
  "model": "qwen-plus",
  "minScore": 60,
  "maxScore": 99,
  "chromePath": ""
}

使用

cd scripts
npm install
npx playwright install chromium
npx playwright test auto-grade.spec.js --headed

获取 API Key

访问阿里云百炼
开通通义千问服务
创建 API Key

安全使用建议

This skill appears coherent with its stated purpose, but before installing/running: 1) Review the entire script (the provided file was truncated in the review input) to confirm no hidden network endpoints or extra behavior. 2) Be aware that student answers will be sent to Aliyun's AIGC endpoint — check privacy, legal, and institutional policies before transmitting student data. 3) Do not put production/high‑privilege Chaoxing or Aliyun credentials in plain config.json; use a dedicated account and API key with minimal permissions and monitor usage to control cost. 4) Run the tool in a controlled/local environment (not on a shared machine), inspect network traffic if possible, and rotate/revoke keys after testing. 5) Confirm that automating grading and auto-submission does not violate Chaoxing/your institution's terms of service or academic integrity rules.

功能分析

Type: OpenClaw Skill Name: chaoxing-auto-grade Version: 1.0.0 The skill automates Chaoxing assignment grading using Playwright and the Qwen AI API. It is classified as suspicious due to security vulnerabilities, primarily a prompt injection risk in `scripts/auto-grade.spec.js` where student answers are directly concatenated into the AI scoring prompt without sanitization. Additionally, the script stores credentials in plaintext within `config.json` and disables HTTPS certificate validation (`ignoreHTTPSErrors: true`). While the code appears to function as described and lacks evidence of intentional malice or data exfiltration, these flaws allow for potential manipulation of grading results and reduced transport security.

能力评估

✓ Purpose & Capability

Name/description match the code and instructions: Playwright automation to log into Chaoxing, find unreviewed assignments, call an Aliyun '通义千问' text-generation endpoint to score subjective answers, and submit scores. Required inputs (Chaoxing username/password and an Aliyun API key) are consistent with that purpose.

ℹ Instruction Scope

SKILL.md and README instruct the user to edit config.json with username/password/apiKey and to run npm install + playwright install then run the provided Playwright test. The runtime instructions and script read only config.json and interact with the Chaoxing web UI and the Aliyun endpoint. Important privacy-related behavior: student answers are sent to dashscope.aliyuncs.com (Aliyun) for scoring — this is expected for the feature but is a privacy/data‑exfiltration risk the user should be aware of.

ℹ Install Mechanism

No packaged install spec in skill metadata (instruction-only), but the README instructs running npm install and 'npx playwright install chromium', which will download packages and browser binaries to disk. This is standard for Playwright-based tools but means code and browser binaries will be written/executed locally; review dependencies and run in a controlled environment.

ℹ Credentials

The skill asks only for the Chaoxing username/password and an Aliyun API key via config.json (declared in skill.json). Those credentials are proportional to the described function. Caveats: credentials and student answers are stored/transmitted in plaintext (config.json and HTTPS requests using the API key); users should avoid reusing high‑privilege credentials and consider separating/testing with throwaway accounts and scoped API keys.

✓ Persistence & Privilege

Skill is not marked always:true and does not request persistent system privileges. It runs as a user-initiated Playwright script; it does not modify other skill configs or attempt broad system changes in the inspected portion.

如何使用

确保已安装 OpenClaw（本地或 Docker 部署）
在对话框中输入安装命令：/install chaoxing-auto-grade
安装完成后，直接呼叫该 Skill 的名称或使用 /chaoxing-auto-grade 触发
根据 Skill 的参数说明提供必要输入，即可获得结构化输出

版本历史

v1.0.0

Chaoxing Auto Grade 1.0.0 – Initial Release - Automates grading of subjective questions on Chaoxing using Playwright and Tongyi Qianwen AI. - Supports auto login, course selection, locating ungraded assignments, AI scoring, and auto submission. - Handles multiple assignments with pagination. - Configuration via simple config.json for credentials and API details.

元数据

Slug chaoxing-auto-grade

版本 1.0.0

许可证 MIT-0

累计安装 0

当前安装数 0

历史版本数 1

常见问题