This skill assists with reading technical books through chapter-by-chapter analysis, comprehension testing, and persistent note-taking. Use this skill when the user wants to read and deeply understand a technical book (PDF/EPUB format), needs structured reading assistance across multiple sessions, or wants to track progress and maintain organized reading notes. Triggers include requests to start reading a book, analyze chapters, save reading notes, track terminology, or schedule review sessions.
This skill enables systematic reading and comprehension of technical books through chapter-by-chapter analysis, interactive learning assistance, and persistent note-taking. Designed for multi-session learning workflows, it helps readers extract key insights, validate understanding, and maintain organized knowledge bases from technical literature.
The skill operates in three reading stages (pre-reading, during-reading, post-reading) and maintains persistent files for continuous learning across sessions.
When the user wants to start reading a new book, follow this initialization sequence:
Ask the user to provide:
Use the Read tool to load the book file and identify the table of contents:
Initialize the output directory with this structure:
<output-directory>/
├── book-metadata.json
├── chapters/
├── glossary.md
├── cross-chapter-analysis.md
└── review-schedule.json
Create the book metadata file with this structure:
{
"title": "Book Title",
"author": "Author Name",
"file_path": "/path/to/book.pdf",
"total_chapters": 12,
"table_of_contents": [
{
"chapter": 1,
"title": "Chapter Title",
"page_start": 15,
"page_end": 42,
"status": "not_started",
"last_accessed": null,
"comprehension_score": null
}
],
"reading_progress": {
"chapters_completed": 0,
"current_chapter": null,
"last_session": null,
"total_reading_time_minutes": 0
},
"created_at": "2025-01-20T10:30:00Z"
}
Create empty placeholder files:
glossary.md: "# Technical Glossary\n\nTerms will be added as you read"cross-chapter-analysis.md: "# Cross-Chapter Analysis\n\n## Recurring Themes\n\n## Conceptual Connections\n\n## Evolution of Ideas"review-schedule.json: {"reviews": []}Inform the user that the book has been initialized and show:
When the user indicates they want to analyze a chapter (e.g., "I'm about to read Chapter 3" or "Let's analyze Chapter 1"):
Identify what stage of reading assistance is needed:
Execute the appropriate reading stage workflow (detailed in next section)
After each interaction, update the relevant files:
Goal: Prepare the reader for effective chapter comprehension by previewing key concepts.
Process:
Output format:
## Chapter X Pre-Reading Brief
### Focus Areas
1. [Key concept 1] - This section introduces...
2. [Key concept 2] - Pay attention to...
3. [Key concept 3] - Understanding this will help...
### Key Terms to Watch For
- **Term 1**: You'll encounter this in the context of...
- **Term 2**: This builds on the concept from Chapter Y...
### Chapter Roadmap
- Introduction (pages X-Y): Sets up the problem
- Core Content (pages Y-Z): Presents three main solutions
- Examples (pages Z-W): Practical applications
### Prerequisite Connections
This chapter builds on concepts from:
- Chapter X: [concept]
- Chapter Y: [concept]
Goal: Enhance comprehension through content expansion and divergent thinking.
Process:
Interaction style:
Goal: Validate comprehension, create comprehensive chapter notes, and schedule reviews.
Process:
Generate a comprehensive chapter note file using the 9-section format (see Chapter Note Format section below). Save this as chapters/chapter-XX-[title-slug].md
Conduct a comprehension check dialogue:
## Comprehension Check - Chapter X
Let's verify your understanding of the key concepts.
### Question 1: [Core concept]
[Thought-provoking question requiring synthesis]
*[After user answers, provide feedback]*
✓ Correct! You've grasped...
or
↻ Not quite. The chapter emphasized that...
### Question 2: [Application]
[Question about practical application]
### Question 3: [Connection]
How does [concept X] relate to [concept Y] from Chapter Z?
Ask 3-5 questions based on the chapter's core teachings. Evaluate user responses and provide constructive feedback.
Based on the user's answers, assign a comprehension score (0-100):
Update book-metadata.json:
Add review entries to review-schedule.json based on the spaced repetition algorithm (see Spaced Repetition section)
Extract new technical terms from the chapter and add to glossary.md with definitions
If this chapter introduces themes or concepts that connect to previous chapters, update cross-chapter-analysis.md
Each chapter analysis file (chapters/chapter-XX-[title-slug].md) follows this 9-section structure:
# Chapter X: [Chapter Title]
## 1. Chapter Metadata
- **Chapter Number**: X
- **Chapter Title**: [Title]
- **Page Range**: XX-YY
- **Date Completed**: YYYY-MM-DD
- **Comprehension Score**: XX/100
## 2. Key Quotes
> "Quote 1 - most powerful or central idea from the chapter"
> *(Page XX)*
> "Quote 2 - thought-provoking statement"
> *(Page YY)*
[4-8 quotes total - capture verbatim with page numbers]
## 3. Main Stories / Examples
**Example 1: [Title]**
- [Brief summary of story/anecdote]
- **Moral/Meaning**: [Key takeaway]
**Example 2: [Title]**
- [Summary]
- **Moral/Meaning**: [Takeaway]
## 4. Chapter Summary
[A clear, concise paragraph (4-6 sentences) summarizing the entire chapter. Capture the main arc: what problem is introduced, what solutions/concepts are presented, and what conclusions are reached.]
## 5. Core Teachings
The main ideas, arguments, or lessons from this chapter:
1. **[Teaching 1]**: [Explanation]
2. **[Teaching 2]**: [Explanation]
3. **[Teaching 3]**: [Explanation]
[Continue as needed]
## 6. Actionable Lessons
Practical lessons or advice that can be applied:
- **[Lesson 1]**: [How to apply]
- **[Lesson 2]**: [How to apply]
- **[Lesson 3]**: [How to apply]
## 7. Mindset / Philosophical Insights
Deeper reflections, shifts in thinking, or philosophical takeaways:
- [Insight 1]
- [Insight 2]
- [Insight 3]
## 8. Memorable Metaphors & Analogies
**[Metaphor 1]**: [Original metaphor from the book]
- **Meaning**: [What it illustrates]
**[Metaphor 2]**: [Another comparison]
- **Meaning**: [Explanation]
## 9. Questions for Reflection
1. [Thought-provoking question related to core concept]
2. [Question about application or implications]
3. [Question connecting to broader context]
4. [Question encouraging critical thinking]
5. [Question about personal application]
---
## Reading Notes
[Optional section for any additional observations, personal insights, or questions that arose during reading]
Writing Guidelines:
As the user progresses through multiple chapters, maintain cross-chapter-analysis.md to track:
Identify concepts, principles, or topics that appear across multiple chapters:
## Recurring Themes
### Theme: [Name]
- **Chapters**: 1, 3, 5, 7
- **Evolution**: How this theme develops across chapters
- **Key Insight**: What the recurring emphasis reveals
### Theme: [Name]
- **Chapters**: 2, 4, 6
- **Evolution**: ...
Map relationships between concepts from different chapters:
## Conceptual Connections
### Connection: [Concept A] ↔ [Concept B]
- **Concept A** (Chapter X): [Brief description]
- **Concept B** (Chapter Y): [Brief description]
- **Relationship**: How they relate, dependencies, or contrast
Track how the author builds arguments progressively:
## Evolution of Ideas
### Idea: [Core Argument]
- **Chapter 1**: Foundation - [What's established]
- **Chapter 3**: Expansion - [How it's built upon]
- **Chapter 5**: Application - [Practical implementation]
- **Chapter 8**: Advanced - [Complex extensions]
Update frequency: Add to cross-chapter analysis after completing each chapter, but only when genuine connections exist. Not every chapter will contribute to every theme.
Maintain glossary.md as an alphabetically organized reference of technical terms:
# Technical Glossary
## A
**[Term]**
- **Definition**: [Clear, concise explanation]
- **Chapter Introduced**: X
- **Context**: [How it's used in the book]
- **Related Terms**: [term1], [term2]
## B
**[Term]**
- **Definition**: ...
Process:
Schedule chapter reviews using a spaced repetition algorithm to combat the forgetting curve.
Base review schedule on comprehension score:
High comprehension (90-100):
Good comprehension (70-89):
Moderate comprehension (50-69):
Low comprehension (<50):
{
"reviews": [
{
"chapter": 1,
"chapter_title": "Introduction to Neural Networks",
"completed_date": "2025-01-15",
"comprehension_score": 85,
"reviews": [
{
"review_number": 1,
"scheduled_date": "2025-01-18",
"completed": true,
"completed_date": "2025-01-18",
"retention_score": 90
},
{
"review_number": 2,
"scheduled_date": "2025-01-29",
"completed": false,
"completed_date": null,
"retention_score": null
}
]
}
]
}
When a review is due or user requests to review a chapter:
chapters/Provide progress visibility when requested:
When user asks "What's my progress?" or "How far have I gotten?":
## Reading Progress: [Book Title]
**Completion**: X/Y chapters (ZZ%)
**Chapters Completed**: 1, 2, 3, 5
**Current Chapter**: 6
**Chapters Remaining**: 7, 8, 9, 10, 11, 12
**Average Comprehension**: XX/100
**Total Reading Time**: XX hours
**Upcoming Reviews**:
- Chapter 1: Due in 2 days
- Chapter 3: Due in 5 days
**Current Streak**: Read X chapters in the last Y days
Provide a visual status overview:
## Chapter Status
1. Introduction ✓ (Score: 85)
2. Foundations ✓ (Score: 92)
3. Core Concepts ✓ (Score: 78)
4. Advanced Topics ⏸ (In progress)
5. Applications ○ (Not started)
...
To maintain continuity across sessions, track and persist:
Keep in working memory during active reading:
When user returns to reading, check book-metadata.json to determine:
Greet returning users with context:
Welcome back! You're reading "[Book Title]".
**Last session**: Chapter X completed on [date]
**Current progress**: X/Y chapters (ZZ%)
**Suggested next**:
- Continue to Chapter Y
- Review Chapter A (scheduled for today)
After each meaningful interaction, update persistent files:
book-metadata.json: Update timestamps, current chapter, status changesglossary.md: When new terms are identified (batch updates preferable)cross-chapter-analysis.md: When completing a chapter that reveals connectionsreview-schedule.json: When completing a chapter or finishing a reviewUser: "I want to start reading 'Deep Learning' by Goodfellow. The PDF is at ~/Books/deep-learning.pdf"
Response:
User: "I'm about to read Chapter 3"
Response:
User: "I'm reading the section on backpropagation. Can you expand on the chain rule application?"
Response:
User: "I finished Chapter 3"
Response:
User: "What reviews are due?"
Response:
This skill uses references to maintain lean instructions while providing detailed specifications:
Contains detailed templates and schemas for all persistent file formats, including:
Provides in-depth explanation of:
Note: These reference files are loaded into context only when needed for detailed formatting questions or troubleshooting. The main SKILL.md provides sufficient guidance for typical usage.