Introduction

1. Why bother
2. The running vault
3. How to read this book

Part I: The Vault

A Plain Folder of Markdown

1. The five folders
2. Why plain text beats a database here
3. Git for your vault
4. A minimal Obsidian setup
5. Filling in note properties automatically
6. Use it with any AI tool
7. Hands-on: create your vault

Projects, Hub Notes and Session Notes

1. The hub note
2. The session note
3. Why Mistakes and Decisions get headings
4. Not only for code
5. Closing a session
6. Let the AI file its own sessions
7. Hands-on: your first project

Small Notes That Link Together

1. One idea per note
2. Why small notes stay findable
3. Linking over filing
4. From Inbox to Resources
5. Granularity in practice
6. Hands-on: break up a long document

Note Properties Your AI Can Rely On

1. The six properties
2. Who reads it
3. Quote values that contain colons
4. Summary: the field that does the most work
5. Linting
6. Hands-on: lint your vault

Part II: The Graph

Your Vault Is Too Big to Read

1. Three costs of reading everything
2. Retrieval, not reading
3. What this book is not
4. Hands-on: price your own vault

Turning Your Notes into a Graph

1. Getting set up
2. Extracting a vault
3. What a node is
4. How the book’s graph was built
5. How conventions become nodes
6. Hands-on: your first graph

Edges You Can Trust

1. The relation vocabulary
2. Confidence tags
3. Measure your own ratio
4. A bad edge, caught
5. Filtering by trust
6. Hands-on: audit your edges

Communities and Surprising Connections

1. How the graph groups your notes
2. Reading the communities of second-brain
3. The notes everything connects to
4. Surprising connections
5. Hands-on: name your communities

The Three Outputs

1. The graph data file
2. The interactive graph map
3. The written graph report
4. Keep generated files out of the vault
5. Exports we don’t use
6. Hands-on: compare degree with the report

Part III: Retrieval and Routing

One Home per Fact

1. Why copies rot
2. The three homes
3. Pointers, not copies
4. Story is allowed to repeat; status is not
5. Drift
6. Hands-on: find the second homes

Routing Your AI to the Right Notes

1. Let the folder tell your AI where to look
2. The routing table of second-brain
3. What a routing file can hold
4. Letting the assistant read it
5. Testing a routing table
6. Hands-on: write your routing table

Cards That Point, Not Tell

1. Anatomy of a card
2. Status: one home, many readers
3. Keep it short on purpose
4. Updating a card
5. Tracing a status question
6. Hands-on: three cards

Asking the Graph Without AI

1. The graph’s building blocks
2. A small recall script
3. Two ways to ask the graph
4. Payoffs: status, decisions and note properties
5. Wiring it in
6. What it costs
7. Limits
8. Hands-on: build and compare

Part IV: Living With the System

Incremental Updates

1. What changes, and how often
2. How the graph remembers what changed
3. Watching and hooks
4. When a full rebuild is fine
5. Stale edges after deletions
6. Measuring a full run against an incremental one
7. Hands-on: keep the graph fresh

Maintenance and Anti-Patterns

1. The six anti-patterns
2. A checkup script for your vault
3. What it found in second-brain
4. The weekly review
5. Archiving and pruning
6. Health metrics
7. Hands-on: build your doctor

Putting It Together

1. The whole picture
2. One project, start to finish
3. Measuring the savings
4. Teams
5. Portability across tools and models
6. In the wild: kdeps
7. Hands-on: the complete loop
8. Closing

Appendix A: Note Properties Reference

Appendix B: Hub Note Template

Appendix C: Session Note Template

Appendix D: Project Card Template

Appendix E: Routing File Template