quiz-generator
Compare original and translation side by side
🇺🇸
Original
English🇨🇳
Translation
ChineseQuiz Generator for Intelligent Textbooks
智能教材测验生成器
Version: 0.3
版本: 0.3
Overview
概述
- For each markdown chapter, generate interactive multiple-choice quizzes for textbook chapters with quality distractor analysis.
- Generate quality reports in markdown format.
- Update mkdocs.yml navigation to include quizzes and reports.
- NEW in v0.3: Support parallel execution for faster generation across multiple chapters.
- 针对每个Markdown章节,生成带有高质量干扰项分析的交互式多项选择题测验。
- 生成Markdown格式的质量报告。
- 更新mkdocs.yml导航栏,添加测验和报告入口。
- v0.3新增功能: 支持并行执行,可快速生成多章节测验。
Purpose
用途
This skill automates quiz creation for intelligent textbooks by analyzing chapter content to generate contextually relevant multiple-choice questions. Each quiz is aligned to specific concepts from the learning graph, distributed across Bloom's Taxonomy cognitive levels, and formatted using mkdocs-material question admonition format with upper-alpha (A, B, C, D) answer choices. The skill ensures quality distractors, balanced answer distribution, and comprehensive explanations for educational value.
本Skill通过分析章节内容,自动为智能教材创建上下文相关的多项选择题。每个测验都与学习图谱中的特定概念对齐,按Bloom's Taxonomy认知层级分布,并使用mkdocs-material的问答提示框格式,采用大写字母(A、B、C、D)作为选项标识。该Skill确保干扰项质量、答案分布均衡,并提供具有教育价值的完整解析。
When to Use This Skill
使用时机
Use this skill after:
- Chapter content has been generated or written (1000+ words per chapter)
- Learning graph exists with concept dependencies
- Glossary is available (recommended for terminology questions)
Trigger this skill when:
- Creating quizzes for new chapters
- Updating quizzes after content revisions
- Building comprehensive quiz bank for entire textbook
- Exporting quiz data for LMS or chatbot integration
The skill can run:
- Parallel mode (default): Multiple chapters processed concurrently for faster generation
- Sequential mode: Chapter by chapter processing
- Single chapter mode: Generate quiz for one specific chapter
请在以下条件满足后使用本Skill:
- 章节内容已生成或撰写完成(每章不少于1000词)
- 学习图谱已存在,且包含概念依赖关系
- 术语表可用(推荐用于术语类题目)
在以下场景触发本Skill:
- 为新章节创建测验
- 内容修订后更新测验
- 为整本教材构建完整的测验题库
- 导出测验数据用于LMS(学习管理系统)或聊天机器人集成
本Skill支持三种运行模式:
- 并行模式(默认):同时处理多个章节,生成速度更快
- 串行模式:逐章处理
- 单章节模式:仅为指定的某一章节生成测验
Execution Modes
执行模式
Parallel Mode (Default for 4+ chapters)
并行模式(4章及以上默认使用)
When generating quizzes for 4 or more chapters, use parallel execution:
| Aspect | Sequential | Parallel |
|---|---|---|
| Agents | 1 | 4-6 concurrent |
| Wall-clock time | ~10 minutes (23 chapters) | ~2-3 minutes |
| Total tokens | Same | Same |
当为4章或更多章节生成测验时,使用并行执行:
| 维度 | 串行模式 | 并行模式 |
|---|---|---|
| Agent数量 | 1 | 4-6个并发 |
| 实际耗时 | ~10分钟(23章) | ~2-3分钟 |
| 总Token消耗 | 相同 | 相同 |
Sequential Mode
串行模式
Use for:
- Fewer than 4 chapters
- Debugging or troubleshooting
- When explicit sequential processing is requested
适用于:
- 章节数量少于4章
- 调试或故障排查
- 用户明确要求串行处理
Single Chapter Mode
单章节模式
Use for:
- Updating one quiz after content revision
- Testing quiz format before batch generation
适用于:
- 内容修订后更新单个测验
- 批量生成前测试测验格式
Workflow
工作流程
Phase 1: Setup (Sequential)
阶段1:准备(串行执行)
This phase runs once before any quiz generation, reading shared context that all agents will need.
本阶段在测验生成前仅执行一次,读取所有Agent都需要的共享上下文。
Step 1.1: Capture Start Time
步骤1.1:记录开始时间
bash
date "+%Y-%m-%d %H:%M:%S"Log the start time for the session report.
bash
date "+%Y-%m-%d %H:%M:%S"记录会话报告的开始时间。
Step 1.2: Indicate Skill Running
步骤1.2:通知Skill运行状态
Notify the user: "Quiz Generator Skill v0.3 running in [parallel/sequential] mode."
向用户发送通知:"Quiz Generator Skill v0.3 正在[并行/串行]模式下运行。"
Step 1.3: Read Shared Context
步骤1.3:读取共享上下文
Read and cache these files for all agents:
-
Course Description ()
docs/course-description.md- Extract target audience and reading level
- Note Bloom's Taxonomy learning outcomes
-
Learning Graph (or similar)
docs/learning-graph/learning-graph.csv- Load concept list with dependencies
- Calculate concept centrality for prioritization
-
Glossary ()
docs/glossary.md- Load term definitions for terminology questions
- Note which concepts have glossary entries
-
Chapter List (scandirectory)
docs/chapters/- Enumerate all chapter directories
- Count words per chapter for readiness assessment
读取并缓存以下文件供所有Agent使用:
-
课程描述()
docs/course-description.md- 提取目标受众和阅读水平
- 记录Bloom's Taxonomy学习目标
-
学习图谱(或类似格式)
docs/learning-graph/learning-graph.csv- 加载包含依赖关系的概念列表
- 计算概念中心性以确定优先级
-
术语表()
docs/glossary.md- 加载术语定义用于术语类题目
- 记录哪些概念在术语表中有定义
-
章节列表(扫描目录)
docs/chapters/- 枚举所有章节目录
- 统计每章的词数以评估内容就绪度
Step 1.4: Assess Content Readiness
步骤1.4:评估内容就绪度
Calculate content readiness score (1-100) for each target chapter:
Quality Checks:
为每个目标章节计算内容就绪度得分(1-100):
质量检查项:
1. Chapter word count:
1. 章节词数:
- 2000+ words = excellent (20 pts)
- 1000-1999 words = good (15 pts)
- 500-999 words = basic (10 pts)
- <500 words = insufficient (5 pts)
- 2000词及以上:优秀(20分)
- 1000-1999词:良好(15分)
- 500-999词:基础(10分)
- 不足500词:不合格(5分)
2. Example coverage:
2. 示例覆盖度:
- 60%+ concepts with examples = excellent (20 pts)
- 40-59% = good (15 pts)
- 20-39% = basic (10 pts)
- <20% = insufficient (5 pts)
- 60%及以上概念配有示例:优秀(20分)
- 40-59%:良好(15分)
- 20-39%:基础(10分)
- 不足20%:不合格(5分)
3. Glossary coverage:
3. 术语表覆盖度:
- 80%+ chapter concepts defined = excellent (20 pts)
- 60-79% = good (15 pts)
- 40-59% = basic (10 pts)
- <40% = insufficient (5 pts)
- 80%及以上章节概念在术语表中有定义:优秀(20分)
- 60-79%:良好(15分)
- 40-59%:基础(10分)
- 不足40%:不合格(5分)
4. Concept clarity:
4. 概念清晰度:
- Clear explanations for all concepts (20 pts)
- Most concepts clear (15 pts)
- Some unclear concepts (10 pts)
- Many unclear concepts (5 pts)
- 所有概念解释清晰(20分)
- 大部分概念解释清晰(15分)
- 部分概念解释模糊(10分)
- 大量概念解释模糊(5分)
5. Learning graph alignment:
5. 学习图谱对齐度:
- All chapter concepts mapped (20 pts)
- Most mapped (15 pts)
- Some mapped (10 pts)
- Few mapped (5 pts)
Content Readiness Ranges:
- 90-100: Rich content, excellent quiz quality possible
- 70-89: Good content, solid quiz possible
- 50-69: Basic content, limited quiz possible
- Below 50: Insufficient content for quality quiz
User Dialog Triggers:
- Score < 60: Ask "Chapter [X] has limited content ([N] words). Generate shorter quiz or skip?"
- No glossary: Ask "No glossary found. Definition questions will be limited. Proceed?"
- Concept gaps: Ask "[N] concepts in chapter not in learning graph. Continue with available concepts?"
- No learning outcomes: Ask "No Bloom's Taxonomy outcomes in course description. Use default distribution?"
- 所有章节概念已映射到学习图谱(20分)
- 大部分概念已映射(15分)
- 部分概念已映射(10分)
- 少数概念已映射(5分)
内容就绪度区间:
- 90-100:内容丰富,可生成高质量测验
- 70-89:内容良好,可生成可靠测验
- 50-69:内容基础,仅能生成有限测验
- 低于50:内容不足,无法生成高质量测验
用户交互触发条件:
- 得分<60:询问“章节[X]内容有限(共[N]词)。是否生成短版测验或跳过该章节?”
- 无术语表:询问“未找到术语表。定义类题目将受限。是否继续?”
- 概念缺失:询问“章节中有[N]个概念未在学习图谱中。是否使用现有概念继续?”
- 无学习目标:询问“课程描述中未包含Bloom's Taxonomy学习目标。是否使用默认分布?”
Step 1.5: Plan Chapter Batches (Parallel Mode)
步骤1.5:规划章节批次(并行模式)
Divide chapters into batches for parallel processing:
Batch Size Guidelines:
- 4-8 chapters: 2 agents (2-4 chapters each)
- 9-15 chapters: 3-4 agents (3-4 chapters each)
- 16-24 chapters: 4-6 agents (4-5 chapters each)
- 25+ chapters: 5-6 agents (5-6 chapters each)
Example for 23 chapters:
Agent 1: Chapters 1-4 (Foundations)
Agent 2: Chapters 5-8 (Limits & Continuity)
Agent 3: Chapters 9-12 (Derivative Rules)
Agent 4: Chapters 13-16 (Applications)
Agent 5: Chapters 17-20 (Analysis)
Agent 6: Chapters 21-23 (Integration)将章节划分为批次以进行并行处理:
批次大小指南:
- 4-8章:2个Agent(每个处理2-4章)
- 9-15章:3-4个Agent(每个处理3-4章)
- 16-24章:4-6个Agent(每个处理4-5章)
- 25章及以上:5-6个Agent(每个处理5-6章)
23章示例:
Agent 1:第1-4章(基础篇)
Agent 2:第5-8章(极限与连续性)
Agent 3:第9-12章(导数规则)
Agent 4:第13-16章(应用篇)
Agent 5:第17-20章(分析篇)
Agent 6:第21-23章(积分篇)Phase 2: Quiz Generation (Parallel or Sequential)
阶段2:测验生成(并行或串行执行)
Parallel Execution
并行执行
Spawn multiple Task agents simultaneously using the Task tool. Each agent receives:
- Shared context (course info, glossary terms, Bloom's targets)
- Assigned chapters (specific chapter directories)
- Quiz format template (the standard format from this skill)
- Output instructions (write quiz.md to each chapter directory)
Agent Prompt Template:
You are generating quizzes for an intelligent textbook. Generate quizzes for
the following chapters.
COURSE CONTEXT:
- Course: [course name]
- Target audience: [audience]
- Reading level: [level]
BLOOM'S TAXONOMY TARGETS:
- Introductory chapters (1-3): 40% Remember, 40% Understand, 15% Apply, 5% Analyze
- Intermediate chapters (4-N): 25% Remember, 30% Understand, 30% Apply, 15% Analyze
- Advanced chapters: 15% Remember, 20% Understand, 25% Apply, 25% Analyze, 10% Evaluate, 5% Create
CHAPTERS TO PROCESS:
[List specific chapter directories with full paths]
FOR EACH CHAPTER:
1. Read the chapter content at the index.md file
2. Identify the key concepts covered in that chapter
3. Generate exactly 10 questions following the format below
4. Ensure answer balance: A (2-3), B (2-3), C (2-3), D (2-3)
5. Write the quiz to docs/chapters/[chapter-dir]/quiz.md
QUIZ FORMAT - Each question MUST follow this exact format:使用Task工具同时启动多个Task Agent。每个Agent将收到:
- 共享上下文(课程信息、术语表术语、Bloom's Taxonomy目标)
- 分配的章节(特定章节目录)
- 测验格式模板(本Skill提供的标准格式)
- 输出说明(将测验写入各章节目录的quiz.md文件)
Agent提示模板:
你正在为智能教材生成测验。请为以下章节生成测验。
课程上下文:
- 课程:[课程名称]
- 目标受众:[受众]
- 阅读水平:[水平]
BLOOM'S TAXONOMY目标:
- 入门章节(1-3章):40% 记忆、40% 理解、15% 应用、5% 分析
- 中级章节(4-N章):25% 记忆、30% 理解、30% 应用、15% 分析
- 高级章节:15% 记忆、20% 理解、25% 应用、25% 分析、10% 评估、5% 创造
待处理章节:
[列出带完整路径的特定章节目录]
针对每章:
1. 读取index.md文件中的章节内容
2. 识别章节涵盖的核心概念
3. 严格按照以下格式生成10道题目
4. 确保答案分布均衡:A(2-3题)、B(2-3题)、C(2-3题)、D(2-3题)
5. 将测验写入docs/chapters/[chapter-dir]/quiz.md
测验格式 - 每道题必须严格遵循以下格式:[N]. [Question text ending with ?]
[题号]. [问题文本,以问号结尾]
<div class="upper-alpha" markdown>
1. [Option A text]
2. [Option B text]
3. [Option C text]
4. [Option D text]
</div>
??? question "Show Answer"
The correct answer is [LETTER]. [Explanation 50-100 words]
**Concept Tested:** [Concept Name]QUIZ FILE STRUCTURE:
<div class="upper-alpha" markdown>
1. [选项A文本]
2. [选项B文本]
3. [选项C文本]
4. [选项D文本]
</div>
??? question "显示答案"
正确答案是**[字母]**。[50-100词的解析]
**测试概念:** [概念名称]测验文件结构:
Quiz: [Chapter Title]
测验:[章节标题]
Test your understanding of [topic] with these review questions.
[Questions 1-10 following the format above]
REPORT when done:
- Chapter name
- Number of questions
- Bloom's distribution (R:#, U:#, Ap:#, An:#)
- Answer distribution (A:#, B:#, C:#, D:#)
**Launching Parallel Agents:**
Use the Task tool with multiple invocations in a SINGLE message to run agents in parallel:
```markdown
[Call Task tool for Agent 1: Chapters 1-4]
[Call Task tool for Agent 2: Chapters 5-8]
[Call Task tool for Agent 3: Chapters 9-12]
[Call Task tool for Agent 4: Chapters 13-16]
[Call Task tool for Agent 5: Chapters 17-20]
[Call Task tool for Agent 6: Chapters 21-23]IMPORTANT: All Task tool calls MUST be in a single message to execute in parallel. If sent in separate messages, they will run sequentially.
通过以下复习题测试你对[主题]的理解程度。
[上述格式的1-10题]
完成后报告:
- 章节名称
- 题目数量
- Bloom's分类分布(记忆:#、理解:#、应用:#、分析:#)
- 答案分布(A:#、B:#、C:#、D:#)
**启动并行Agent:**
在同一条消息中多次调用Task工具,以实现Agent并行运行:
```markdown
[为Agent 1调用Task工具:处理第1-4章]
[为Agent 2调用Task工具:处理第5-8章]
[为Agent 3调用Task工具:处理第9-12章]
[为Agent 4调用Task工具:处理第13-16章]
[为Agent 5调用Task工具:处理第17-20章]
[为Agent 6调用Task工具:处理第21-23章]重要提示: 所有Task工具调用必须放在同一条消息中才能实现并行执行。如果分开发送,将默认串行运行。
Sequential Execution
串行执行
For sequential mode or fewer than 4 chapters, process each chapter one at a time following Steps 2-8 below.
对于串行模式或章节数少于4章的情况,按照以下步骤2-8逐章处理。
Phase 2 Steps (Per Chapter - used by agents or sequential mode)
阶段2步骤(每章执行 - Agent或串行模式通用)
Step 2: Determine Target Distribution
步骤2:确定目标分布
Based on chapter type (introductory, intermediate, advanced), set target Bloom's Taxonomy distribution:
Introductory Chapters (typically chapters 1-3):
- 40% Remember
- 40% Understand
- 15% Apply
- 5% Analyze
- 0% Evaluate
- 0% Create
Intermediate Chapters:
- 25% Remember
- 30% Understand
- 30% Apply
- 15% Analyze
- 0% Evaluate
- 0% Create
Advanced Chapters:
- 15% Remember
- 20% Understand
- 25% Apply
- 25% Analyze
- 10% Evaluate
- 5% Create
Determine chapter type by:
- Position in textbook (first 3 chapters = introductory)
- Concept centrality in learning graph (high centrality = advanced)
- Explicit markers in chapter metadata
- User specification
Target question count: 8-12 per chapter (default: 10)
根据章节类型(入门、中级、高级),设置Bloom's Taxonomy目标分布:
入门章节(通常为1-3章):
- 40% 记忆
- 40% 理解
- 15% 应用
- 5% 分析
- 0% 评估
- 0% 创造
中级章节:
- 25% 记忆
- 30% 理解
- 30% 应用
- 15% 分析
- 0% 评估
- 0% 创造
高级章节:
- 15% 记忆
- 20% 理解
- 25% 应用
- 25% 分析
- 10% 评估
- 5% 创造
通过以下方式确定章节类型:
- 在教材中的位置(前3章=入门)
- 学习图谱中的概念中心性(中心性高=高级)
- 章节元数据中的明确标记
- 用户指定
目标题量:每章8-12题(默认10题)
Step 3: Identify Concepts to Test
步骤3:确定待测试概念
Analyze chapter content and learning graph to prioritize concepts:
Priority 1 (Must Test):
- High-centrality concepts in learning graph
- Concepts mentioned in chapter title or introduction
- Concepts with dedicated sections
- Key terms emphasized in bold or glossary links
Priority 2 (Should Test):
- Supporting concepts with substantial explanation
- Concepts with examples
- Prerequisites reviewed in chapter
- Concepts from learning objectives
Priority 3 (May Test):
- Peripheral concepts mentioned briefly
- Related concepts for context
- Future topics previewed
Aim for 80%+ coverage of Priority 1 concepts.
分析章节内容和学习图谱,确定概念优先级:
优先级1(必须测试):
- 学习图谱中的高中心性概念
- 章节标题或引言中提及的概念
- 有专门小节讲解的概念
- 加粗或带有术语表链接的关键术语
优先级2(建议测试):
- 有详细解释的支撑性概念
- 配有示例的概念
- 章节中回顾的前置知识
- 学习目标中列出的概念
优先级3(可选测试):
- 简要提及的边缘概念
- 用于上下文补充的相关概念
- 预告的未来主题
目标:覆盖80%以上的优先级1概念
Step 4: Generate Questions by Bloom's Level
步骤4:按Bloom's层级生成题目
For each concept selected for testing, generate question at appropriate Bloom's level following target distribution.
IMPORTANT FORMATTING REQUIREMENT:
All questions MUST use the mkdocs-material question admonition format with upper-alpha list styling:
markdown
undefined为每个选定的概念生成对应Bloom's层级的题目,符合目标分布要求。
重要格式要求:
所有题目必须使用mkdocs-material的问答提示框格式,并采用大写字母列表样式:
markdown
undefined1. What is the primary purpose of a learning graph?
1. 学习图谱的主要用途是什么?
<div class="upper-alpha" markdown>
1. To create visual decorations for textbooks
2. To map prerequisite relationships between concepts
3. To generate random quiz questions
4. To organize files in a directory structure
</div>
??? question "Show Answer"
The correct answer is B. A learning graph is a directed graph that maps prerequisite relationships between concepts, showing which concepts must be learned before others. This ensures proper scaffolding in educational content.
**Concept Tested:** Learning Graph
**See:** [Learning Graph Concept](../concepts/learning-graph.md)
**Formatting Rules:**
1. Use level-4 header (####) with question number
2. Write question as complete sentence ending with ?
3. Use `<div class="upper-alpha" markdown>` wrapper
4. Write 4 answer options as numbered list (1, 2, 3, 4)
5. Use `??? question "Show Answer"` admonition
6. Indent answer content with 4 spaces
7. Start with "The correct answer is **[LETTER]**."
8. Include concept name and link to source
9. Maintain blank line before and after div
**Question Writing Guidelines:**
**Remember Level:**
- Ask for definitions from glossary
- Test fact recall
- Identify terminology
- Example: "What is the definition of [concept]?"
**Understand Level:**
- Ask for explanations
- Test comprehension of relationships
- Compare/contrast concepts
- Example: "Which best describes the relationship between [A] and [B]?"
**Apply Level:**
- Present scenarios requiring concept application
- Test problem-solving using learned methods
- Example: "Given [scenario], which approach would you use?"
**Analyze Level:**
- Ask to identify patterns or causes
- Test ability to break down concepts
- Example: "What is the underlying reason for [phenomenon]?"
**Evaluate Level:**
- Ask for judgments based on criteria
- Test critical thinking
- Example: "Which approach would be most effective for [goal]?"
**Create Level:**
- Ask to design solutions
- Test synthesis of concepts
- Example: "How would you design a [system] that [requirements]?"<div class="upper-alpha" markdown>
1. 为教材创建视觉装饰
2. 映射概念之间的前置依赖关系
3. 生成随机测验题目
4. 组织目录中的文件结构
</div>
??? question "显示答案"
正确答案是B。学习图谱是一种有向图,用于映射概念之间的前置依赖关系,展示哪些概念需要先学习。这确保了教育内容的合理递进。
**测试概念:** 学习图谱
**参考:** [学习图谱概念](../concepts/learning-graph.md)
**格式规则:**
1. 使用4级标题(####),后跟题号
2. 问题为完整句子,以问号结尾
3. 使用`<div class="upper-alpha" markdown>`包裹选项
4. 4个选项采用编号列表(1、2、3、4)
5. 使用`??? question "显示答案"`提示框
6. 答案内容缩进4个空格
7. 以“正确答案是**[字母]**。”开头
8. 包含概念名称和来源链接
9. 在div前后保留空行
**题目编写指南:**
**记忆层级:**
- 询问术语表中的定义
- 测试事实回忆
- 识别术语
- 示例:“[概念]的定义是什么?”
**理解层级:**
- 询问解释
- 测试对关系的理解
- 对比/比较概念
- 示例:“以下哪项最能描述[A]和[B]之间的关系?”
**应用层级:**
- 呈现需要应用概念的场景
- 测试使用所学方法解决问题的能力
- 示例:“给定[场景],你会使用哪种方法?”
**分析层级:**
- 要求识别模式或原因
- 测试拆解概念的能力
- 示例:“[现象]的根本原因是什么?”
**评估层级:**
- 要求基于标准做出判断
- 测试批判性思维
- 示例:“哪种方法最能实现[目标]?”
**创造层级:**
- 要求设计解决方案
- 测试概念综合能力
- 示例:“你会如何设计一个满足[需求]的[系统]?”Step 5: Write Quality Distractors
步骤5:编写高质量干扰项
For each incorrect answer option (distractors), ensure:
Plausibility:
- Sounds reasonable to someone who hasn't learned the material
- Uses related terminology
- Avoids obviously wrong answers
- Similar length to correct answer
Educational Value:
- Addresses common misconceptions
- Tests understanding of related concepts
- Discriminates between levels of knowledge
- Not trick questions or word games
Common Distractor Patterns:
- Partial truth (correct in different context)
- Reversal (opposite of correct answer)
- Similar terminology (related but distinct concept)
- Common error (typical student mistake)
Avoid:
- "All of the above" or "None of the above"
- Jokes or nonsensical options
- Grammatically inconsistent options
- Answers that overlap or both could be correct
对于每个错误选项(干扰项),需满足以下要求:
合理性:
- 对未学习过该内容的人来说听起来合理
- 使用相关术语
- 避免明显错误的答案
- 长度与正确答案相近
教育价值:
- 针对常见误解
- 测试对相关概念的理解
- 区分不同知识水平
- 不是文字游戏或陷阱题
常见干扰项模式:
- 部分正确(在其他语境下成立)
- 反向表述(与正确答案相反)
- 相似术语(相关但不同的概念)
- 常见错误(学生易犯的典型错误)
避免:
- “以上所有”或“以上皆非”选项
- 玩笑或无意义选项
- 语法不一致的选项
- 重叠或都可能正确的选项
Step 6: Write Explanations
步骤6:编写解析
For each question, write explanation that:
Confirms Correct Answer:
- State clearly: "The correct answer is [LETTER]."
- Explain why this answer is correct
- Reference chapter content or concept definition
- Target: 50-100 words
Teaches (Optional but Recommended):
- Briefly explain why distractors are incorrect
- Clarify common misconceptions
- Provide additional context
- Link to chapter section for more detail
Example Explanation:
The correct answer is **B**. A learning graph is a directed graph that maps
prerequisite relationships between concepts. Option A is incorrect because
learning graphs serve a structural purpose, not decorative. Option C is
incorrect because quiz generation is not the primary purpose. Option D
confuses learning graphs with file systems.
**Concept Tested:** Learning Graph
**See:** [Learning Graph Concept](../concepts/learning-graph.md#definition)为每道题编写解析,需包含:
确认正确答案:
- 明确说明:“正确答案是**[字母]**。”
- 解释该答案正确的原因
- 参考章节内容或概念定义
- 目标长度:50-100词
教学价值(可选但推荐):
- 简要解释干扰项错误的原因
- 澄清常见误解
- 提供额外上下文
- 链接到章节中的对应部分
解析示例:
正确答案是**B**。学习图谱是一种有向图,用于映射概念之间的前置依赖关系。选项A错误,因为学习图谱的作用是结构化内容而非装饰。选项C错误,因为生成测验并非其主要用途。选项D混淆了学习图谱与文件系统。
**测试概念:** 学习图谱
**参考:** [学习图谱概念](../concepts/learning-graph.md#definition)Step 7: Ensure Answer Balance
步骤7:确保答案分布均衡
Check that correct answers are distributed evenly across A, B, C, D:
Target Distribution:
- A: 25% (±5%)
- B: 25% (±5%)
- C: 25% (±5%)
- D: 25% (±5%)
Avoid Patterns:
- All C's in a row
- Alternating A-B-A-B
- Predictable sequences
- Position bias (first/last always correct)
Randomization Strategy:
- Generate random sequence before writing quiz
- Shuffle for each question
- Verify distribution after completion
- Adjust if imbalanced
检查正确答案在A、B、C、D之间的分布是否均匀:
目标分布:
- A:25%(±5%)
- B:25%(±5%)
- C:25%(±5%)
- D:25%(±5%)
避免模式:
- 连续多题答案为C
- 交替出现A-B-A-B
- 可预测的序列
- 位置偏见(首选项或末选项总是正确答案)
随机化策略:
- 编写测验前生成随机序列
- 为每道题打乱选项顺序
- 完成后验证分布情况
- 若分布不均则调整
Step 8: Create Quiz File
步骤8:创建测验文件
Generate quiz file with proper structure:
Separate Quiz File ():
docs/chapters/[chapter-name]/quiz.mdmarkdown
undefined生成结构规范的测验文件:
独立测验文件():
docs/chapters/[chapter-name]/quiz.mdmarkdown
undefinedQuiz: [Chapter Name]
测验:[章节名称]
Test your understanding of [chapter topic] with these questions.
通过以下题目测试你对[章节主题]的理解程度。
1. [Question text]?
1. [问题文本]?
<div class="upper-alpha" markdown>
1. [Option 1]
2. [Option 2]
3. [Option 3]
4. [Option 4]
</div>
??? question "Show Answer"
The correct answer is [LETTER]. [Explanation]
**Concept Tested:** [Concept Name]<div class="upper-alpha" markdown>
1. [选项1]
2. [选项2]
3. [选项3]
4. [选项4]
</div>
??? question "显示答案"
正确答案是**[字母]**。[解析]
**测试概念:** [概念名称]2. [Question text]?
2. [问题文本]?
[Continue for all questions...]
**Formatting Requirements:**
- Use horizontal rules (---) between questions
- Number questions sequentially (1, 2, 3...)
- Maintain consistent spacing
- Ensure all markdown renders correctly[按上述格式继续编写所有题目...]
**格式要求:**
- 题目之间使用水平线(---)分隔
- 题目按顺序编号(1、2、3...)
- 保持一致的间距
- 确保所有Markdown内容可正确渲染Phase 3: Aggregation (Sequential, after parallel agents complete)
阶段3:结果聚合(串行执行,在所有并行Agent完成后进行)
After all parallel agents complete, aggregate results.
所有并行Agent完成后,汇总结果。
Step 9: Collect Agent Results
步骤9:收集Agent结果
Wait for all Task agents to complete. Collect from each:
- List of quiz files created
- Per-chapter statistics (questions, Bloom's distribution, answer balance)
- Any errors or issues encountered
等待所有Task Agent完成,收集每个Agent的以下信息:
- 创建的测验文件列表
- 每章统计数据(题目数量、Bloom's分布、答案均衡性)
- 遇到的任何错误或问题
Step 10: Generate Metadata Files (Optional)
步骤10:生成元数据文件(可选)
Create for each chapter:
docs/learning-graph/quizzes/[chapter-name]-quiz-metadata.jsonjson
{
"chapter": "Chapter Name",
"chapter_file": "docs/chapters/chapter-name/index.md",
"quiz_file": "docs/chapters/chapter-name/quiz.md",
"generated_date": "YYYY-MM-DD",
"total_questions": 10,
"content_readiness_score": 85,
"overall_quality_score": 78,
"questions": [
{
"id": "ch1-q001",
"number": 1,
"question_text": "What is the primary purpose of a learning graph?",
"correct_answer": "B",
"bloom_level": "Understand",
"difficulty": "medium",
"concept_tested": "Learning Graph",
"source_link": "../concepts/learning-graph.md",
"distractor_quality": 0.85,
"explanation_word_count": 67
}
],
"answer_distribution": {
"A": 2,
"B": 3,
"C": 3,
"D": 2
},
"bloom_distribution": {
"Remember": 2,
"Understand": 4,
"Apply": 3,
"Analyze": 1,
"Evaluate": 0,
"Create": 0
},
"concept_coverage": {
"total_concepts": 12,
"tested_concepts": 10,
"coverage_percentage": 83
}
}为每章创建:
docs/learning-graph/quizzes/[chapter-name]-quiz-metadata.jsonjson
{
"chapter": "章节名称",
"chapter_file": "docs/chapters/chapter-name/index.md",
"quiz_file": "docs/chapters/chapter-name/quiz.md",
"generated_date": "YYYY-MM-DD",
"total_questions": 10,
"content_readiness_score": 85,
"overall_quality_score": 78,
"questions": [
{
"id": "ch1-q001",
"number": 1,
"question_text": "学习图谱的主要用途是什么?",
"correct_answer": "B",
"bloom_level": "理解",
"difficulty": "中等",
"concept_tested": "学习图谱",
"source_link": "../concepts/learning-graph.md",
"distractor_quality": 0.85,
"explanation_word_count": 67
}
],
"answer_distribution": {
"A": 2,
"B": 3,
"C": 3,
"D": 2
},
"bloom_distribution": {
"Remember": 2,
"Understand": 4,
"Apply": 3,
"Analyze": 1,
"Evaluate": 0,
"Create": 0
},
"concept_coverage": {
"total_concepts": 12,
"tested_concepts": 10,
"coverage_percentage": 83
}
}Step 11: Generate Quiz Bank (Aggregate)
步骤11:生成测验题库(聚合)
Create or update with all questions:
docs/learning-graph/quiz-bank.jsonjson
{
"textbook_title": "Building Intelligent Textbooks",
"generated_date": "YYYY-MM-DD",
"total_chapters": 20,
"total_questions": 187,
"questions": [
{
"id": "ch1-q001",
"chapter": "Introduction to Learning Graphs",
"question_text": "What is the primary purpose of a learning graph?",
"options": {
"A": "To create visual decorations for textbooks",
"B": "To map prerequisite relationships between concepts",
"C": "To generate random quiz questions",
"D": "To organize files in a directory structure"
},
"correct_answer": "B",
"explanation": "A learning graph is a directed graph...",
"bloom_level": "Understand",
"difficulty": "medium",
"concept": "Learning Graph",
"chapter_file": "docs/concepts/learning-graph.md",
"source_section": "#definition",
"tags": ["graph", "prerequisites", "scaffolding"]
}
]
}Use Cases for Quiz Bank:
- LMS export (Moodle, Canvas, Blackboard XML)
- Quiz randomization (select subset)
- Alternative quiz versions
- Chatbot integration (practice questions)
- Study app integration
创建或更新,包含所有题目:
docs/learning-graph/quiz-bank.jsonjson
{
"textbook_title": "智能教材构建",
"generated_date": "YYYY-MM-DD",
"total_chapters": 20,
"total_questions": 187,
"questions": [
{
"id": "ch1-q001",
"chapter": "学习图谱简介",
"question_text": "学习图谱的主要用途是什么?",
"options": {
"A": "为教材创建视觉装饰",
"B": "映射概念之间的前置依赖关系",
"C": "生成随机测验题目",
"D": "组织目录中的文件结构"
},
"correct_answer": "B",
"explanation": "学习图谱是一种有向图...",
"bloom_level": "理解",
"difficulty": "中等",
"concept": "学习图谱",
"chapter_file": "docs/concepts/learning-graph.md",
"source_section": "#definition",
"tags": ["graph", "prerequisites", "scaffolding"]
}
]
}测验题库用途:
- 导出到LMS(Moodle、Canvas、Blackboard XML格式)
- 测验随机化(选择题目子集)
- 生成替代版本测验
- 聊天机器人集成(练习题)
- 学习应用集成
Step 12: Generate Quality Report
步骤12:生成质量报告
Create :
docs/learning-graph/quiz-generation-report.mdmarkdown
undefined创建:
docs/learning-graph/quiz-generation-report.mdmarkdown
undefinedQuiz Generation Quality Report
测验生成质量报告
Generated: YYYY-MM-DD
Execution Mode: Parallel (6 agents)
Wall-clock Time: X minutes Y seconds
生成时间:YYYY-MM-DD
执行模式:并行(6个Agent)
实际耗时:X分Y秒
Overall Statistics
总体统计
- Total Chapters: 20
- Total Questions: 187
- Avg Questions per Chapter: 9.4
- Overall Quality Score: 76/100
- 总章节数: 20
- 总题目数: 187
- 每章平均题目数: 9.4
- 总体质量得分: 76/100
Execution Summary (Parallel Mode)
执行总结(并行模式)
| Agent | Chapters | Questions | Time |
|---|---|---|---|
| Agent 1 | 1-4 | 40 | 45s |
| Agent 2 | 5-8 | 40 | 52s |
| Agent 3 | 9-12 | 40 | 48s |
| Agent 4 | 13-16 | 40 | 51s |
| Agent 5 | 17-20 | 40 | 47s |
| Agent 6 | 21-23 | 27 | 38s |
| Agent | 处理章节 | 题目数量 | 耗时 |
|---|---|---|---|
| Agent 1 | 1-4章 | 40 | 45秒 |
| Agent 2 | 5-8章 | 40 | 52秒 |
| Agent 3 | 9-12章 | 40 | 48秒 |
| Agent 4 | 13-16章 | 40 | 51秒 |
| Agent 5 | 17-20章 | 40 | 47秒 |
| Agent 6 | 21-23章 | 27 | 38秒 |
Per-Chapter Summary
每章总结
| Chapter | Questions | Quality Score | Bloom's Score | Coverage |
|---|---|---|---|---|
| Ch 1: Introduction | 10 | 82/100 | 24/25 | 83% |
| Ch 2: Learning Graphs | 12 | 78/100 | 22/25 | 90% |
| ... | ... | ... | ... | ... |
| 章节 | 题目数量 | 质量得分 | Bloom's得分 | 概念覆盖度 |
|---|---|---|---|---|
| 第1章:简介 | 10 | 82/100 | 24/25 | 83% |
| 第2章:学习图谱 | 12 | 78/100 | 22/25 | 90% |
| ... | ... | ... | ... | ... |
Bloom's Taxonomy Distribution (Overall)
总体Bloom's Taxonomy分布
| Level | Actual | Target | Deviation |
|---|---|---|---|
| Remember | 22% | 25% | -3% ✓ |
| Understand | 28% | 30% | -2% ✓ |
| Apply | 27% | 25% | +2% ✓ |
| Analyze | 18% | 15% | +3% ✓ |
| Evaluate | 4% | 4% | 0% ✓ |
| Create | 1% | 1% | 0% ✓ |
Bloom's Distribution Score: 24/25 (excellent)
| 层级 | 实际占比 | 目标占比 | 偏差 |
|---|---|---|---|
| 记忆 | 22% | 25% | -3% ✓ |
| 理解 | 28% | 30% | -2% ✓ |
| 应用 | 27% | 25% | +2% ✓ |
| 分析 | 18% | 15% | +3% ✓ |
| 评估 | 4% | 4% | 0% ✓ |
| 创造 | 1% | 1% | 0% ✓ |
Bloom's分布得分: 24/25(优秀)
Answer Balance (Overall)
总体答案分布
- A: 24% (45/187)
- B: 26% (49/187)
- C: 25% (47/187)
- D: 25% (46/187)
Answer Balance Score: 15/15 (perfect distribution)
- A:24%(45/187)
- B:26%(49/187)
- C:25%(47/187)
- D:25%(46/187)
答案均衡性得分: 15/15(完美分布)
Recommendations
建议
[Include recommendations based on aggregated data]
undefined[基于聚合数据给出建议]
undefinedStep 13: Validate Quality
步骤13:质量验证
Perform comprehensive validation across all generated quizzes:
1. No Ambiguity:
- Each question has exactly one defensible correct answer
- Question stem is clear and complete
- No grammatical errors
2. Distractor Quality:
- All wrong answers are plausible
- Distractors test understanding, not just guessing
- Similar length and grammatical structure
- No overlapping answers
3. Grammar & Clarity:
- Professional writing throughout
- Consistent verb tense
- Proper punctuation
- No typos
4. Answer Balance:
- Correct answers distributed across A, B, C, D
- Within 20-30% per option (target: 25%)
- No predictable patterns
5. Bloom's Distribution:
- Matches target for chapter type
- Within ±15% acceptable
- Progressive difficulty through quiz
6. Concept Coverage:
- 75%+ of major concepts tested
- Important concepts have multiple questions
- No over-testing trivial concepts
7. No Duplicates:
- Unique questions across all quizzes
- No near-duplicates (>80% similar)
8. Explanation Quality:
- All questions have explanations
- Explanations teach, not just confirm
- 50-100 words target
- Reference chapter sections
9. Link Validation:
- All source links point to existing content
- Use section anchors where appropriate
- Links render correctly
- Do not place links in quiz.md files that do not work
- Do not use link labels to sections that do not exist
10. Bias Check:
- No cultural bias
- No gender bias
- No assumptions about background
- Accessible language
Success Criteria:
- Overall quality score > 70/100
- 8-12 questions per chapter
- Bloom's distribution within ±15% of target
- 75%+ concept coverage
- Answer balance within 20-30% per option
- 100% questions have explanations
- No duplicate questions
- All links valid
对所有生成的测验进行全面验证:
1. 无歧义:
- 每道题仅有一个明确正确的答案
- 题干清晰完整
- 无语法错误
2. 干扰项质量:
- 所有错误选项都具有合理性
- 干扰项测试理解能力,而非单纯猜测
- 长度和语法结构与正确答案相似
- 无重叠选项
3. 语法与清晰度:
- 全程使用专业书面语
- 动词时态一致
- 标点符号正确
- 无拼写错误
4. 答案均衡性:
- 正确答案均匀分布在A、B、C、D之间
- 每个选项占比在20-30%之间(目标25%)
- 无可预测的模式
5. Bloom's分布:
- 符合对应章节类型的目标分布
- 偏差在±15%以内为可接受
- 测验难度逐步提升
6. 概念覆盖度:
- 75%以上的主要概念已测试
- 重要概念配有多道题目
- 未过度测试琐碎概念
7. 无重复:
- 所有测验中的题目均唯一
- 无高度相似的题目(相似度>80%)
8. 解析质量:
- 所有题目都有解析
- 解析具有教学价值,而非仅确认答案
- 长度在50-100词之间
- 参考章节对应部分
9. 链接验证:
- 所有来源链接指向有效内容
- 适当使用章节锚点
- 链接可正确渲染
- 不在quiz.md文件中添加无效链接
- 不使用指向不存在章节的链接标签
10. 偏见检查:
- 无文化偏见
- 无性别偏见
- 不对用户背景做出假设
- 使用易懂的语言
成功标准:
- 总体质量得分>70/100
- 每章8-12题
- Bloom's分布与目标偏差在±15%以内
- 概念覆盖度≥75%
- 每个选项占比在20-30%之间
- 100%的题目配有解析
- 无重复题目
- 所有链接有效
Step 14: Update Site Navigation
步骤14:更新站点导航
Update to include quizzes in each chapter directory:
mkdocs.ymlyml
nav:
...
- Chapters:
- Overview: chapters/index.md
- 1. Introduction to AI and Intelligent Textbooks:
- Content: chapters/01-intro-ai-intelligent-textbooks/index.md
- Quiz: chapters/01-intro-ai-intelligent-textbooks/quiz.md
- 2. Getting Started with Claude and Skills:
- Content: chapters/02-getting-started-claude-skills/index.md
- Quiz: chapters/02-getting-started-claude-skills/quiz.md
- 3. Course Design and Educational Theory:
- Content: chapters/03-course-design-educational-theory/index.md
- Quiz: chapters/03-course-design-educational-theory/quiz.mdNote that the string "Chapter" should not be placed in the main chapter content label that points to the index.md file.
Also update to include quiz quality reports:
mkdocs.ymlyml
nav:
...
Learning Graph:
...
Quiz Generation Report: learning-graph/quiz-generation-report.md更新,在每个章节目录中添加测验入口:
mkdocs.ymlyml
nav:
...
- 章节:
- 概述:chapters/index.md
- 1. AI与智能教材简介:
- 内容:chapters/01-intro-ai-intelligent-textbooks/index.md
- 测验:chapters/01-intro-ai-intelligent-textbooks/quiz.md
- 2. Claude与Skill入门:
- 内容:chapters/02-getting-started-claude-skills/index.md
- 测验:chapters/02-getting-started-claude-skills/quiz.md
- 3. 课程设计与教育理论:
- 内容:chapters/03-course-design-educational-theory/index.md
- 测验:chapters/03-course-design-educational-theory/quiz.md注意:指向index.md的主章节标签中不应包含“章节”字样。
同时更新,添加测验质量报告入口:
mkdocs.ymlyml
nav:
...
学习图谱:
...
测验生成报告:learning-graph/quiz-generation-report.mdStep 15: Capture End Time and Write Session Log
步骤15:记录结束时间并写入会话日志
Capture the end time:
bash
date "+%Y-%m-%d %H:%M:%S"Export the session information to :
logs/quiz-generator-YYYY-MM-DD.mdmarkdown
undefined记录结束时间:
bash
date "+%Y-%m-%d %H:%M:%S"将会话信息导出到:
logs/quiz-generator-YYYY-MM-DD.mdmarkdown
undefinedQuiz Generator Session Log
测验生成器会话日志
Skill Version: 0.3
Date: YYYY-MM-DD
Execution Mode: Parallel (6 agents)
Skill版本: 0.3
日期: YYYY-MM-DD
执行模式: 并行(6个Agent)
Timing
时间统计
| Metric | Value |
|---|---|
| Start Time | YYYY-MM-DD HH:MM:SS |
| End Time | YYYY-MM-DD HH:MM:SS |
| Elapsed Time | X minutes Y seconds |
| 指标 | 数值 |
|---|---|
| 开始时间 | YYYY-MM-DD HH:MM:SS |
| 结束时间 | YYYY-MM-DD HH:MM:SS |
| 耗时 | X分Y秒 |
Token Usage
Token使用量
| Phase | Estimated Tokens |
|---|---|
| Setup (shared context) | ~15,000 |
| Agent 1 (Ch 1-4) | ~25,000 |
| Agent 2 (Ch 5-8) | ~25,000 |
| ... | ... |
| Aggregation | ~5,000 |
| Total | ~160,000 |
| 阶段 | 预估Token数 |
|---|---|
| 准备阶段(共享上下文) | ~15,000 |
| Agent 1(第1-4章) | ~25,000 |
| Agent 2(第5-8章) | ~25,000 |
| ... | ... |
| 聚合阶段 | ~5,000 |
| 总计 | ~160,000 |
Results
结果
- Total chapters: N
- Total questions: N × 10
- Quality score: XX/100
- All quizzes written successfully: Yes/No
- 总章节数:N
- 总题目数:N × 10
- 质量得分:XX/100
- 所有测验创建成功:是/否
Files Created
创建的文件
[List all quiz.md files and report files]
undefined[列出所有quiz.md文件和报告文件]
undefinedStep 16: Notify User
步骤16:通知用户
Notify the user:
"Quiz Generator v0.3 complete!
- Mode: Parallel (6 agents)
- Elapsed time: X minutes Y seconds
- Chapters processed: 23
- Questions generated: 230
- Quality score: 82/100
The site navigation in has been updated to include Content/Quiz links for each chapter and the quiz generation report in the learning-graph section.
mkdocs.ymlSession logged to "
logs/quiz-generator-YYYY-MM-DD.md向用户发送通知:
"测验生成器v0.3已完成!
- 模式: 并行(6个Agent)
- 耗时: X分Y秒
- 处理章节数: 23
- 生成题目数: 230
- 质量得分: 82/100
已更新中的站点导航,为每个章节添加“内容/测验”链接,并在学习图谱部分添加测验生成报告入口。
mkdocs.yml会话日志已保存到"
logs/quiz-generator-YYYY-MM-DD.mdQuestion Format Reference
题目格式参考
Complete Example with All Elements
包含所有元素的完整示例
markdown
undefinedmarkdown
undefined3. Given a course with 50 concepts, what is the most important factor in organizing the learning graph?
3. 假设一门课程包含50个概念,组织学习图谱时最重要的因素是什么?
<div class="upper-alpha" markdown>
1. Alphabetical order of concept names
2. Prerequisite relationships between concepts
3. The length of concept definitions
4. The visual appearance of the graph diagram
</div>
??? question "Show Answer"
The correct answer is B. Prerequisite relationships are the most important factor because they determine the order in which concepts must be learned. A learning graph maps these dependencies to ensure students learn foundational concepts before advanced ones. Alphabetical order (A) and visual appearance (D) are organizational preferences, not educational requirements. Definition length (C) does not affect concept sequencing.
**Concept Tested:** Learning Graph Structure
**See:** [Learning Graph](../concepts/learning-graph.md#prerequisites)undefined<div class="upper-alpha" markdown>
1. 概念名称的字母顺序
2. 概念之间的前置依赖关系
3. 概念定义的长度
4. 图谱的视觉外观
</div>
??? question "显示答案"
正确答案是B。前置依赖关系是最重要的因素,因为它决定了概念的学习顺序。学习图谱通过映射这些依赖关系,确保学生先学习基础概念,再学习高级概念。字母顺序(A)和视觉外观(D)是组织偏好,而非教育需求。定义长度(C)不影响概念排序。
**测试概念:** 学习图谱结构
**参考:** [学习图谱](../concepts/learning-graph.md#prerequisites)undefinedFormatting Checklist
格式检查清单
- Level-4 header with question number
- Complete sentence ending with ?
- wrapper
<div class="upper-alpha" markdown> - Numbered list (1, 2, 3, 4) for options
- Closing tag
</div> - admonition
??? question "Show Answer" - 4-space indentation in answer block
- "The correct answer is [LETTER]." statement
- Explanation (50-100 words)
- Concept Tested: label
- See: link with proper path
- Blank lines before and after div
- 使用4级标题,后跟题号
- 问题为完整句子,以问号结尾
- 使用包裹选项
<div class="upper-alpha" markdown> - 选项采用编号列表(1、2、3、4)
- 使用提示框
??? question "显示答案" - 答案内容缩进4个空格
- 以“正确答案是**[字母]**。”开头
- 包含概念名称和来源链接
- 在div前后保留空行
Common Pitfalls to Avoid
常见误区
Format Errors:
- ❌ Forgetting wrapper
<div class="upper-alpha" markdown> - ❌ Using letters (A, B, C, D) instead of numbers in list
- ❌ Incorrect indentation in answer block
- ❌ Missing closing tag
</div>
Question Quality:
- ❌ Ambiguous questions with multiple correct answers
- ❌ "All of the above" or "None of the above" options
- ❌ Trick questions or word games
- ❌ Questions testing trivial facts
Distractor Quality:
- ❌ Obviously wrong answers
- ❌ Joke options or nonsense
- ❌ Distractors much longer/shorter than correct answer
- ❌ Options that overlap or contradict
Explanation Quality:
- ❌ Just restating the question
- ❌ No teaching value
- ❌ Missing or broken links
- ❌ Too brief (< 30 words) or too long (> 150 words)
Parallel Execution:
- ❌ Sending Task calls in separate messages (runs sequentially)
- ❌ Not waiting for all agents before aggregation
- ❌ Forgetting to aggregate statistics from all agents
格式错误:
- ❌ 忘记添加包裹
<div class="upper-alpha" markdown> - ❌ 选项使用字母(A、B、C、D)而非数字
- ❌ 答案块缩进错误
- ❌ 缺少闭合标签
</div>
题目质量问题:
- ❌ 歧义题(多个正确答案)
- ❌ 使用“以上所有”或“以上皆非”选项
- ❌ 文字游戏或陷阱题
- ❌ 测试琐碎事实的题目
干扰项质量问题:
- ❌ 明显错误的选项
- ❌ 玩笑或无意义选项
- ❌ 干扰项长度与正确答案差异过大
- ❌ 重叠或矛盾的选项
解析质量问题:
- ❌ 仅重复题干内容
- ❌ 无教学价值
- ❌ 链接缺失或无效
- ❌ 过短(<30词)或过长(>150词)
并行执行问题:
- ❌ 分开发送Task调用(将串行运行)
- ❌ 在聚合前未等待所有Agent完成
- ❌ 忘记汇总所有Agent的统计数据
Output Files Summary
输出文件汇总
Required (Per Chapter):
- Quiz markdown file:
docs/chapters/[chapter-name]/quiz.md
Recommended (Aggregate):
2. - Quality metrics
3. - Session log with timing
docs/learning-graph/quiz-generation-report.mdlogs/quiz-generator-YYYY-MM-DD.mdOptional:
4. - All questions database
5. - Per-chapter metadata
6. Navigation updates to
docs/learning-graph/quiz-bank.jsondocs/learning-graph/quizzes/[chapter-name]-quiz-metadata.jsonmkdocs.yml必填(每章):
- 测验Markdown文件:
docs/chapters/[chapter-name]/quiz.md
推荐(聚合):
2. - 质量指标报告
3. - 包含时间统计的会话日志
docs/learning-graph/quiz-generation-report.mdlogs/quiz-generator-YYYY-MM-DD.md可选:
4. - 全题目数据库
5. - 每章元数据
6. 导航更新
docs/learning-graph/quiz-bank.jsondocs/learning-graph/quizzes/[chapter-name]-quiz-metadata.jsonmkdocs.ymlExample Session
示例会话
Parallel Mode (Default)
并行模式(默认)
User: "Generate quizzes for all chapters"
Claude (using this skill):
- Captures start time
- Notifies: "Quiz Generator Skill v0.3 running in parallel mode."
- Reads shared context (course description, learning graph, glossary)
- Scans chapter directories, finds 23 chapters
- Assesses content readiness (all chapters 2000+ words)
- Plans batches: 6 agents, ~4 chapters each
- Spawns 6 Task agents in a SINGLE message (parallel execution)
- Waits for all agents to complete
- Aggregates results from all agents
- Generates quality report (score: 82/100)
- Updates mkdocs.yml navigation
- Captures end time
- Writes session log
- Reports: "Quiz Generator v0.3 complete! Mode: Parallel. Time: 2m 45s. Questions: 230. Quality: 82/100."
用户: "为所有章节生成测验"
Claude(使用本Skill):
- 记录开始时间
- 发送通知:"测验生成器Skill v0.3正在并行模式下运行。"
- 读取共享上下文(课程描述、学习图谱、术语表)
- 扫描章节目录,发现23章
- 评估内容就绪度(所有章节均超过2000词)
- 规划批次:6个Agent,每个处理约4章
- 在同一条消息中启动6个Task Agent(并行执行)
- 等待所有Agent完成
- 汇总所有Agent的结果
- 生成质量报告(得分:82/100)
- 更新mkdocs.yml导航
- 记录结束时间
- 写入会话日志
- 发送报告:"测验生成器v0.3已完成!模式:并行。耗时:2分45秒。生成题目数:230。质量得分:82/100。"
Sequential Mode
串行模式
User: "Generate a quiz for Chapter 3 only"
Claude (using this skill):
- Assesses Chapter 3 content readiness (score: 82/100)
- Determines chapter type: intermediate
- Sets target distribution: 25% Remember, 30% Understand, 30% Apply, 15% Analyze
- Identifies 12 concepts to test (10 priority 1, 2 priority 2)
- Generates 10 questions using question admonition format
- Creates quality distractors
- Ensures answer balance (A: 2, B: 3, C: 3, D: 2)
- Writes explanations without links unless verified
- Writes quiz to
docs/chapters/03-chapter-name/quiz.md - Reports: "Created 10-question quiz for Chapter 3. Quality score: 78/100."
用户: "仅为第3章生成测验"
Claude(使用本Skill):
- 评估第3章内容就绪度(得分:82/100)
- 确定章节类型:中级
- 设置目标分布:25%记忆、30%理解、30%应用、15%分析
- 识别12个待测试概念(10个优先级1,2个优先级2)
- 使用问答提示框格式生成10道题目
- 创建高质量干扰项
- 确保答案分布均衡(A:2、B:3、C:3、D:2)
- 编写解析(仅在链接验证通过后添加)
- 将测验写入
docs/chapters/03-chapter-name/quiz.md - 发送报告:"已为第3章创建10道题的测验。质量得分:78/100。"