--- description: Create book notes with parallel extraction and optional knowledge compounding via spawned Concept/Pattern/Theme notes model: sonnet --- # /book-notes Create structured book notes with parallel extraction of concepts, frameworks, and actionable insights. Uses three agents to extract different dimensions simultaneously, producing a rich reference note. Optionally spawns standalone knowledge nodes for concepts, patterns, and themes worth preserving independently. ## Usage ``` /book-notes "" --author "" [options] ``` ### Parameters | Parameter | Description | Required | |----------------|------------------------------------------------------|----------| | `title` | Book title | Yes | | `--author` | Book author | Yes | | `--input` | Source material type (default: user provides context) | No | | `--isbn` | ISBN for metadata lookup | No | | `--spawn` | Enable knowledge compounding — spawn standalone notes | No | | `--spawn-only` | Skip book note creation, only spawn from existing | No | ## Instructions ### Phase 1: Gather Input Ask the user for their source material: highlights/annotations, chapter summaries, overall impressions, or key quotes. If the user has no notes, offer to create a template they can fill in, or search for the book online to gather context. ### Phase 2: Parallel Extraction — Agent Team Launch three agents simultaneously using the Task tool. **Agent 1: Concept Extractor** (Sonnet) Task: Identify and describe key concepts and ideas - Extract all distinct concepts, theories, and ideas from the input - For each concept: name, definition, significance, related concepts - Identify the book's central thesis or argument - Map how concepts relate to each other - Note which concepts are original vs building on existing work Return: List of concepts with descriptions and relationships **Agent 2: Framework Extractor** (Sonnet) Task: Identify frameworks, models, and methodologies - Extract all structured frameworks, models, and step-by-step processes - For each: name, purpose, steps/components, when to use it - Identify mental models the author uses or promotes - Note practical tools or templates described - Create visual representations where possible (tables, lists) Return: List of frameworks with descriptions and application guidance **Agent 3: Action Extractor** (Sonnet) Task: Extract practical advice and recommendations - Pull out all actionable advice and recommendations - For each: what to do, when to do it, expected outcome - Identify exercises, experiments, or practices suggested - Note warnings or anti-patterns the author highlights - Prioritise actions by impact and ease of implementation Return: Prioritised list of actionable takeaways with context ### Phase 3: Synthesise Book Note Combine agent outputs into a comprehensive reference note: 1. Book metadata and rating 2. One-paragraph summary 3. Key concepts (from Agent 1) 4. Frameworks and models (from Agent 2) 5. Actionable takeaways (from Agent 3) 6. Key quotes 7. Personal reflections If `--spawn` is active, proceed to Phase 4 after writing the book note. ### Phase 4: Knowledge Compounding (Optional) Activated by the `--spawn` flag. Default behaviour (no flag) creates only the book note. #### Step 1: Check for Existing Spawned Nodes (Idempotent Handling) If the book note already exists, read its frontmatter for a `spawnedNodes` field. | Situation | Action | |------------------------------|-------------------------------------------------| | No existing `spawnedNodes` | Proceed with full analysis | | Has `spawnedNodes` | Skip existing, analyse for new candidates only | When running incrementally, append new nodes to `spawnedNodes` rather than replacing the list. #### Step 2: Vault Deduplication Search Before proposing spawn candidates, search existing notes to avoid duplicates. For each potential candidate: 1. **Exact match search** — check for files named `Concept - `, `Pattern - `, or `Theme - ` 2. **Partial match search** — search note contents for the candidate term **Decision Matrix:** | Search Result | Action | |-----------------------------------------------------|-------------------------------------------| | Exact match exists | **Link** to existing note, do not spawn | | Partial coverage (mentioned but no dedicated note) | Offer to spawn or enrich existing | | Related concept exists | Add to `relatedTo`, may still spawn | | No coverage | Add to **spawn candidates** | Report results before proceeding. #### Step 3: Classify and Assess Candidates Using the extracted concepts (from Phase 2 agents), classify each candidate: | Type | Criteria | |-------------|------------------------------------------------------| | **Concept** | "What is X?" — definitions, frameworks, models | | **Pattern** | "How to do X?" — approaches, architectures, solutions | | **Theme** | Cross-cutting concerns, paradigm shifts, strategic directions | For each candidate, determine: | Field | Description | |------------------|-----------------------------------------------------| | `name` | Note title | | `type` | Concept, Pattern, or Theme | | `definition` | 1-2 sentence definition | | `sourceChapters` | Which chapters cover this | | `sourcePrimary` | Is this book THE authoritative source? | | `confidence` | Derived from source quality (see Confidence table) | #### Step 4: Map Inter-Node Relationships Before creating notes, map how candidates relate to each other using typed relationships: | Relationship Type | Meaning | |-------------------------|----------------------------------------| | "component of" | X is a part of Y | | "implements" | X realises Y in practice | | "enables" | X makes Y possible | | "governs" | X provides rules/constraints for Y | | "defines interface for" | X specifies how to interact with Y | This ensures spawned notes link to siblings, not just back to the source book. #### Step 5: Create Spawned Notes Create notes in this order (so earlier notes can be referenced by later ones): 1. Themes first (broadest scope) 2. Concepts second (foundational) 3. Patterns last (apply concepts) For each node: 1. Create file using the appropriate spawned note template (see below) 2. Populate provenance fields: `sourceBook`, `sourceChapters`, `sourcePrimary` 3. Set `confidence` based on the inheritance rules 4. Include `relatedTo` linking to sibling nodes AND the source book #### Step 6: Update Book Note Add the `spawnedNodes` field to the book note frontmatter and append a "Spawned Knowledge" section to the body: ```markdown ## Spawned Knowledge This book is a primary source for the following knowledge notes: ### Themes | Note | Description | |------|-------------| | [[Concept - ]] | | ### Core Concepts | Note | Description | Chapters | |------|-------------|----------| | [[Concept - ]] | | Ch. N | ### Implementation Patterns | Note | Description | Chapters | |------|-------------|----------| | [[Pattern - ]] | | Ch. N | ``` ## Output Format ### Book Note Template ```markdown --- type: Reference title: "Reference - " referenceType: book created: YYYY-MM-DD author: "" isbn: "" yearPublished: YYYY tags: [content/book, domain/relevant-tag] rating: X/5 summary: "" # Knowledge Compounding (populated by --spawn) spawnedNodes: [] --- # > **Author:** | **Published:** YYYY | **Rating:** X/5 ## One-Paragraph Summary ## Key Concepts ## Frameworks and Models ## Actionable Takeaways ## Key Quotes ## Personal Reflections ## Related Reading ``` ### Spawned Concept Template **Filename:** `Concept - .md` Create for "What is X?" knowledge — definitions, frameworks, models. ```markdown --- type: Concept title: "" aliases: [] created: YYYY-MM-DD tags: [] # Source Provenance sourceBook: "[[Reference - ]]" sourceChapters: [N, N] sourcePrimary: true|false # Relationships relatedTo: - "[[Reference - ]]" - "[[Concept - ]]" # Quality summary: "" confidence: high|medium|low --- # > **Definition:** ## Overview <2-3 paragraphs explaining the concept, written to be standalone> ## Key Characteristics - - ## Related Concepts - [[Concept - ]] — ## Sources - [[Reference - ]] — Ch. N (primary|secondary source) ``` ### Spawned Pattern Template **Filename:** `Pattern - .md` Create for "How to do X?" knowledge — approaches, architectures, solutions. ```markdown --- type: Pattern title: "" aliases: [] created: YYYY-MM-DD tags: [] # Source Provenance sourceBook: "[[Reference - ]]" sourceChapters: [N, N] sourcePrimary: true|false # Relationships relatedTo: - "[[Reference - ]]" - "[[Concept - ]]" # Quality summary: "" confidence: high|medium|low --- # > **Intent:** ## Context and Problem ## Solution ## Consequences **Benefits:** / **Trade-offs:** ## Related Patterns - [[Pattern - ]] — ## Sources - [[Reference - ]] — Ch. N ``` ### Spawned Theme Template **Filename:** `Concept - .md` (themes use `conceptType: theme`) Create for cross-cutting concerns, paradigm shifts, strategic directions. ```markdown --- type: Concept title: "" conceptType: theme aliases: [] created: YYYY-MM-DD tags: [] # Source Provenance sourceBook: "[[Reference - ]]" sourceChapters: [N, N] sourcePrimary: true|false # Relationships relatedTo: - "[[Reference - ]]" - "[[Concept - ]]" - "[[Pattern - ]]" # Quality summary: "" confidence: high|medium|low --- # > **Essence:** ## Overview <2-3 paragraphs explaining the theme and its significance> ## Manifestations - [[Concept - ]] — - [[Pattern - ]] — ## Tensions and Trade-offs ## Sources - [[Reference - ]] — Primary exposition ``` ## Knowledge Compounding Reference ### Confidence Inheritance | Source Quality | Criteria | Spawned Confidence | |-----------------------------------------|-------------------------------------------------------|--------------------| | **Authoritative** | Author coined or invented the term | `high` | | **Deep treatment** | Dedicated chapter(s), comprehensive coverage | `high` | | **Substantial** | Multiple sections, detailed explanation | `medium` | | **Referenced** | Mentioned, explained briefly | `medium` | | **Brief mention** | Passing reference, assumes prior knowledge | `low` | ### Typical Spawn Counts | Book Type | Themes | Concepts | Patterns | |--------------------------|--------|----------|----------| | Technical deep-dive | 1-3 | 5-10 | 3-7 | | Survey / "Fundamentals" | 2-4 | 10-20 | 5-10 | | Practical guide | 1-2 | 3-7 | 5-15 | | Business / strategy | 2-4 | 3-5 | 1-3 | ### Spawn Decision Rules **Do spawn** when the term needs standalone explanation, the book is THE primary source, or the concept/pattern will be referenced from multiple notes. **Do not spawn** when an existing note covers it (link instead), the term is too narrow or book-specific, or coverage is too brief to support a useful standalone note. ## Examples ### Example 1: From Highlights ``` /book-notes "Thinking in Systems" --author "Donella Meadows" --input highlights ``` User provides Kindle highlights; agents extract systems thinking concepts, leverage points framework, and actionable mental models. ### Example 2: With Knowledge Compounding ``` /book-notes "Team Topologies" --author "Matthew Skelton" --spawn ``` Creates the book note, then identifies candidates such as Concept - Cognitive Load, Concept - Team API, Pattern - Stream-Aligned Team, and Theme - Team-First Thinking. Searches the vault for existing coverage, presents candidates for approval, and creates standalone notes with provenance tracking. ### Example 3: Spawn from Existing Book Note ``` /book-notes "Designing Data-Intensive Applications" --author "Martin Kleppmann" --spawn-only ``` Reads the existing book note, identifies concepts and patterns not yet spawned, and creates standalone notes without recreating the book note. --- **Invoke with:** `/book-notes "" --author "<author>"` to create structured book notes with parallel extraction, or add `--spawn` to also create standalone knowledge notes