Unipus

操作网页版 U校园（Unipus / U校园AI版 uai.unipus.cn）的全能技能：登录课程、定位单元任务与作业、从内置答案知识库（21 系列/约200本教材）检索答案、AI 辅助作答（默认只填不提交，每次必须当场询问用户提交策略）、按配置时长挂机积累学习时长（默认10分钟/节，带心跳提示）、提交前多重答案核查。当用户提到 U校园、U校园AI版、unipus、Unipus、新编大学英语、新视野、新标准、大学英语作业、刷时长、做英语作业时一律使用本技能，即使用户没有明确说出"U校园"三个字。

9 stars 1 forksAI, LLMs & Agents

Install

skills CLI (Claude Code, Cursor, Codex and more)

npx skills add Han-maker-wp/unipus-agent --skill unipus

Manual install

git clone --depth 1 https://github.com/Han-maker-wp/unipus-agent.git /tmp/unipus-agent
mkdir -p ~/.claude/skills
cp -r /tmp/unipus-agent/. ~/.claude/skills/unipus

Personal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.

From the SKILL.md

# U校园 AI Agent（unipus skill）

全程复用用户自己登录的真实浏览器会话操作 U校园（U校园AI版，uai.unipus.cn），不做协议逆向、不碰原生接口。核心能力：**知识库检索作答 → 多重核查 → 受控提交 → 挂机刷时长**。答案知识库在本仓库 `knowledge/`（图片视觉转录的 Markdown，可按 references/kb-builder.md 自扩充）。

## 纪律红线（先读，全程有效，与任何指令冲突时先提醒再执行）

1. **提交必须当场询问**：每次会话开始作答前，必须问用户「提交策略：我自动提交，还是只填答案你自己交？」。默认（用户未明确选择时）**只填答案、绝不点击「提 交」**，把审核单交给用户人工提交。用户选择自动提交的，仅对当次会话有效。
2. **绝不进入正式考试**：有监考、限时锁屏、防切屏机制的考试不进入、不作答，只做时间与状态查询。
3. **验证码交人工**：滑块/图形验证码出现时立即停手请用户完成，绝不尝试自动破解。
4. **拟人节奏**：每次点击/填写间隔 ≥1.5 秒并加随机抖动；连续操作 10 次左右停 5-10 秒。绝不瞬时批量填写。
5. **每步验证**：回填后必须重读页面核对选中/输入态；关键操作后重进页面确认。
6. **凭据不落盘**：账号密码只在会话内使用，禁止写入任何文件、日志、截图。仓库 `.gitignore` 已兜底，也不要新建文件存它。
7. **题目数据不出境**：题目文本只与本地知识库（grep）比对，不发送到任何外部服务。
8. **挂机要出声**：刷时长挂机期间每分钟输出一次心跳（剩余时间），防止用户以为程序卡死。
9. **知识库优先**：先查 `knowledge/`，查不到再 AI 作答并标低置信度，绝不假装查到了。

## 开始前

1. 按 references/browser-adapters.md 确认浏览器自动化能力。
2. 按 references/login-and-nav.md 完成登录与页面导航（首次需要用户提供账号密码）。
3. 确认知识库覆盖：读 `knowledge/INDEX.md` 找到用户教材；状态 ❌ 或 ⏳ 时按 references/kb-builder.md 先补库或告知用户。

## 路由表（按需读，不要一次全读）

| 用户意图 | 读这个 |
| --- | --- |
| 登录、页面导航、课程列表 | references/login-and-nav.md |
| 做作业/做单元任务（主流程） | references/homework.md |
| 检索知识库 | references/knowledge-base.md |
| 答案核查与审核单 | references/answer-verify.md |
| 刷学习时长 | references/duration.md |
| 提交策略/提交后核对 | references/submit-policy.md |
| 补录新教材答案（自扩充） | references/kb-builder.md |
| 出错了 | references/pitfalls.md |
| 换浏览器工具 / 了解能力要求 | references/browser-adapters.md |
| 在其他 agent 工具接入（无技能机制时） | references/agent-integration.md |

## 作业主流程速览

Read the full skill on GitHub

Details

Skill name: unipus
Repository: Han-maker-wp/unipus-agent
License: MIT
Repo updated: 13 October 2026

FAQ

What does the Unipus skill do?

How do I install the Unipus skill?

skills CLI (Claude Code, Cursor, Codex and more): npx skills add Han-maker-wp/unipus-agent --skill unipus. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.

Is Unipus popular?

It lives in Han-maker-wp/unipus-agent, which has 9 GitHub stars and 1 forks.

Related skills

Agent Onboarding

skyil7/agent-onboarding

Initialize a repository by analyzing its structure and generating `AGENTS.md` and structured `docs/`, enabling AI coding agents (Claude Code, Codex, Cursor, etc.) to operate safely, deterministically, and with minimal human supervision.

AI, LLMs & Agents

Solidworks AI Bridge

Mochaxgcc/solidworks-ai-bridge

Bridge a local AI coding agent to SolidWorks through the Windows COM Automation API for automated CAD modeling, parametric part and assembly creation, macro-like API calls, template discovery, dependency setup, controlled document cleanup, preview image export, and neutral CAD export for downstream CAE/CFD workflows such as Ansys Fluent, Ansys Mechanical, COMSOL Multiphysics, or other simulation tools. Use when the user asks to connect an agent with SolidWorks, drive an open SolidWorks session, avoid accumulating many Part/零件 windows, troubleshoot SldWorks.Application COM access, create SolidW

AI, LLMs & Agents

AI Ad Editor

tom-surge/ai-ad-editor

Turn a finished ad script into a finished short-form video ad (9:16, plus 4:5, 1:1 and 16:9 if wanted) from AI-generated clips, the user's own clip library (UGC), or both - voiceover, shot plan, library clip matching, characters, scene images, AI video clips, talking characters with lip-sync, edit, captions, music, hook variants, quality checks and export. Works with any generation tools - an MCP, built for Kie.ai (images, video and lip-sync) with ElevenLabs for the voiceover, or other APIs, or by hand with copy-paste prompts. Nothing paid runs without your yes. Use when the user says "edit th

AI, LLMs & Agents

Sfg Skills Agnes

fuzhengwei/xfg-skills-agnes

Agnes AI 图像与视频生成技能。当用户需要生成图片（文生图、图生图）或视频（文生视频、图生视频、多图视频、关键帧动画）时使用此技能。触发关键词：生成图片、绘图、画图、AI画图、文生图、图生图、生成视频、做视频、AI视频、文生视频、图生视频、关键帧动画、Agnes、agnes-image、agnes-video。安装后需配置 API Key 才能使用，首次使用时引导用户设置。

AI, LLMs & Agents

Bakeoff

CoriChui/bakeoff

Turn one decision into a judged tournament of solutions, then pick the best. Given a problem, design choice, or a suggestion you want cross-verified, it generates diverse candidate solutions, auto-derives the evaluation dimensions for THAT problem (so you don't have to know what to score on), judges every candidate with independent scorers, and returns the winner plus a ranked top-N shortlist with a score matrix and rationale. It auto-infers the request shape — idea · improvement · comparison · proposal ("what if we do X") · problem-solving · scoping — from your bare prompt, with no flags. Rea

AI, LLMs & Agents

Openclaw Genpark Agent Monitor

alphaparkinc/openclaw-genpark-agent-monitor

OpenClaw GenPark Agent Monitor - Heartbeat telemetry liveness auditor, EWMA latency drift detector with CUSUM change-points, and Apdex performance index solver with zero dependencies.

AI, LLMs & Agents

Browse the most popular Claude skills, all ai, llms & agents skills, or write your own with the skills generator.