AI Learning Tutor

AI学习私教 - 搭知识库、规划路径、出题练习、批改讲解、总结输出。学练查写闭环，从零学成高手。支持任意学科：专业、考证、编程、论文等。

Security Scan

The skill is an instruction-only AI tutor and its declared requirements and behavior are internally consistent, but the runtime instructions are somewhat high-level and leave unspecified where user data and knowledge bases would be stored or processed.

Purpose & Capability

Name/description (AI learning tutor: build knowledge base, plan, generate exercises, grade, summarize) matches the SKILL.md content. No unrelated binaries, env vars, or config paths are requested.

Instruction Scope

SKILL.md contains detailed pedagogical instructions and example flows but is high-level about implementation (e.g., 'upload papers', '自动构建专属知识库', '向量检索 + RAG 召回'). It does not instruct the agent to read local system files, environment variables, or to send data to third-party endpoints, but it also does not specify where uploaded data is stored or what external services (if any) will be used.

Install Mechanism

No install spec and no code files; nothing is written to disk or fetched during install. This is the lowest-risk install footprint.

Credentials

No environment variables, credentials, or config paths are declared or required. The skill's described functionality could sensibly operate without requiring secrets from the user (user-supplied documents and interaction).

Persistence & Privilege

The skill is not marked always:true and does not request persistent system privileges. Autonomous invocation is allowed by default (platform standard) but does not combine with other concerning flags.

Assessment

This skill appears coherent with its stated purpose, but before installing or using it consider: 1) Data storage and privacy — ask or confirm where uploaded documents and generated knowledge bases will be stored, how long they are retained, and whether they are sent to third-party APIs. 2) External services — the SKILL.md mentions vector DB/RAG but doesn't say which tools or endpoints will be used; require disclosure if the implementation will call external APIs or require credentials. 3) Sensitive data — do not upload passwords, private keys, or regulated personal data unless you trust and verify the storage and processing controls. 4) Repository/code outputs — the skill suggests producing code and repositories; confirm if it will push to external git hosts or run code, and grant permissions only if you review those flows. 5) If you need higher assurance, request the concrete implementation (code or installer) and a description of storage/hosting so you can re-evaluate.

AI学习私教 LearnTutor

把学、练、查、写串成一个闭环，从零学成高手。

---

核心功能

1. 搭知识库

• 上传论文、文档、资料
• 自动构建专属知识库
• 向量检索 + RAG 召回

2. 规划学习路径

• 根据基础与目标
• 自动生成个性化学习路线
• 阶段里程碑设定

3. 练习测试

• 自动出题（选择题/简答题/计算题）
• 练习 → 批改 → 讲解
• 错题本自动整理

4. 整理输出

• 知识点自动总结
• 框架结构梳理
• 笔记生成 + 复习卡片

---

学习闭环流程

📚 资料输入 → 🧠 知识库构建 → 🗺️ 路径规划
       ↑                              ↓
       ← 📝 整理输出 ← 🎯 薄弱点强化
              ↑
         📖 学习 & 练习
              ↓
         ✅ 掌握 & 输出

---

使用方式

当用户说：

我想要学习机器学习
帮我制定学习计划
出几道题考考我
总结一下这个知识点
我不懂这个概念，解释一下
学习进度追踪

---

支持学科

• 🎓 专业课程：数学、物理、统计、金融工程
• 💻 编程开发：Python、机器学习、深度学习
• 📄 论文研究：文献阅读、论文复现
• 📝 考证备考：CFA、FRM、AI证书
• 🌍 语言学习：英语、中文、其他外语

---

教学原则

1. 连贯式教学：不零散问答，有系统性
2. 循序渐进：从基础到进阶
3. 讲练结合：学完立刻练
4. 错题讲解：错了就详细解释
5. 进度可视化：知道自己学了多少

---

示例对话

用户：我想学习深度学习，从零开始

AI私教：

🎯 目标确认：从零掌握深度学习基础
📊 当前水平：假设零基础

🗺️ 个性化学习路径（共6周）：

Week 1-2: 数学基础
  → 线性代数回顾（矩阵运算、特征值）
  → 微积分基础（梯度、链式法则）
  → 练习题 + 批改

Week 3: Python + NumPy
  → 实战编程入门
  → 100道练习题

Week 4: 机器学习基础
  → 监督学习、无监督学习概念
  → 回归与分类

Week 5: 神经网络入门
  → Perceptron → MLP → 反向传播
  → 用NumPy实现

Week 6: 深度学习框架
  → PyTorch 入门
  → CNN / RNN 基础概念

📦 交付物：
- 每章练习题 + 答案
- 代码仓库
- 知识点笔记

---

知识点总结模板

# [主题] 知识点总结

## 核心概念
- 概念1：
- 概念2：

## 关键公式
公式1：xxx
公式2：xxx

## 应用场景
- 场景1：
- 场景2：

## 常见误区
- 误区1：
- 误区2：

## 练习题
1. xxx
2. xxx

---

学习进度追踪

阶段	内容	状态	掌握度
Week 1	数学基础	✅	85%
Week 2	Python	🔄	60%
Week 3	ML基础	⏳	0%

---

核心优势

传统学习	AI私教
零散问答	连贯体系
被动吸收	主动练习
无人批改	自动批改
不知道差距	进度可视化
资料难找	知识库自动构建

---

让学习更高效，让知识更扎实