- Cover
- Colophon
- Epigraph
- 1. Introduction
❱
- 1.1. Welcome
- 1.2. What Is Agentic Coding?
- 1.3. What Are Design Patterns?
- 1.4. How to Read This Book
- 1.5. How This Book Writes Itself
- 1.6. Learning Tracks
- 1.7. What's New
- 1.8. Article Map
- 1.9. Meta Report
- 2. Product Judgment and What to Create
❱
- 2.1. Problem
- 2.2. Customer
- 2.3. User
- 2.4. Value Proposition
- 2.5. Competitive Landscape
- 2.6. Differentiation
- 2.7. Beachhead
- 2.8. Go-to-Market
- 2.9. Revenue Model
- 2.10. Monetization
- 2.11. Distribution
- 2.12. Product-Market Fit
- 2.13. Crossing the Chasm
- 2.14. Zero to One
- 2.15. Bottleneck
- 2.16. Roadmap
- 2.17. User Story
- 2.18. Use Case
- 2.19. Build-vs-Don't-Build Judgment
- 3. Intent, Scope, and Decision-Making
❱
- 3.1. Application
- 3.2. Brief
- 3.3. Requirement
- 3.4. Constraint
- 3.5. Acceptance Criteria
- 3.6. Specification
- 3.7. Spec-Driven Development
- 3.8. Design Doc
- 3.9. Tradeoff
- 3.10. Judgment
- 3.11. Taste
- 3.12. Architecture Decision Record
- 3.13. Risk Spike
- 3.14. Programming Language Selection
- 4. Structure and Decomposition
❱
- 4.1. Architecture
- 4.2. Shape
- 4.3. Abstraction
- 4.4. Component
- 4.5. Module
- 4.6. Interface
- 4.7. Consumer
- 4.8. Contract
- 4.9. Boundary
- 4.10. Cohesion
- 4.11. Coupling
- 4.12. Dependency
- 4.13. Composition
- 4.14. Separation of Concerns
- 4.15. Monolith
- 4.16. Decomposition
- 4.17. Task Decomposition
- 4.18. Big Ball of Mud
- 4.19. Spaghetti Code
- 4.20. God Object
- 5. Data, State, and Truth
❱
- 5.1. Data Model
- 5.2. Schema (Database)
- 5.3. Schema (Serialization)
- 5.4. Data Structure
- 5.5. State
- 5.6. Artifact
- 5.7. Source of Truth
- 5.8. DRY (Don't Repeat Yourself)
- 5.9. Copy-Paste Programming
- 5.10. Hard Coding
- 5.11. Primitive Obsession
- 5.12. Data Normalization / Denormalization
- 5.13. Database
- 5.14. CRUD
- 5.15. Consistency
- 5.16. Atomic
- 5.17. Transaction
- 5.18. Serialization
- 5.19. Idempotency
- 5.20. Domain Model
- 5.21. Entity
- 5.22. Value Object
- 5.23. Ubiquitous Language
- 5.24. Naming
- 5.25. Coding Convention
- 5.26. Aggregate
- 5.27. Bounded Context
- 5.28. Anti-Corruption Layer
- 5.29. Business Capability
- 6. Computation and Interaction
❱
- 6.1. Algorithm
- 6.2. Algorithmic Complexity
- 6.3. API
- 6.4. Protocol
- 6.5. Determinism
- 6.6. Side Effect
- 6.7. Concurrency
- 6.8. Event
- 7. Correctness, Testing, and Evolution
❱
- 7.1. Invariant
- 7.2. Test
- 7.3. Test Oracle
- 7.4. Code Coverage
- 7.5. LLM-as-Judge
- 7.6. Static Analysis
- 7.7. Formal Verification
- 7.8. Harness
- 7.9. Fixture
- 7.10. Test Double
- 7.11. Test-Driven Development
- 7.12. Red/Green TDD
- 7.13. Reproduction Test
- 7.14. Mutation Testing
- 7.15. Property-Based Testing
- 7.16. Approval Testing
- 7.17. Refactor
- 7.18. Regression
- 7.19. Test Impact Analysis
- 7.20. Flaky Test Quarantine
- 7.21. Test Pyramid
- 7.22. Smoke Test
- 7.23. Exploratory Testing
- 7.24. Agentic Manual Testing
- 7.25. Consumer-Driven Contract Testing
- 7.26. Observability
- 7.27. Domain-Oriented Observability
- 7.28. Agent Trace
- 7.29. Failure Mode
- 7.30. Failure Response Mode
- 7.31. Layers and Generators
- 7.32. Silent Failure
- 7.33. Fail Fast and Loud
- 7.34. Performance Envelope
- 7.35. Logging
- 7.36. Happy Path
- 7.37. Production-Readiness Cliff
- 7.38. Code Review
- 7.39. Review Bottleneck
- 7.40. Risk-Based Review
- 7.41. Printf Debugging
- 7.42. Scientific Debugging
- 7.43. Metric
- 7.44. Feedback Loop
- 7.45. Service Level Objective
- 7.46. Technical Debt
- 7.47. Greenfield and Brownfield
- 7.48. Strangler Fig
- 7.49. Parallel Change
- 7.50. Deprecation
- 7.51. Evolutionary Modernization
- 7.52. Regenerative Software
- 7.53. Sweep
- 7.54. Backfill
- 8. Security and Trust
❱
- 8.1. Threat Model
- 8.2. Attack Surface
- 8.3. Trust Boundary
- 8.4. Authentication
- 8.5. Authorization
- 8.6. Vulnerability
- 8.7. Least Privilege
- 8.8. Agentic Payments
- 8.9. Secret
- 8.10. Input Validation
- 8.11. Output Encoding
- 8.12. Sandbox
- 8.13. Agent Gateway
- 8.14. Action-Selector
- 8.15. Blast Radius
- 8.16. Prompt Injection
- 8.17. Tool Poisoning
- 8.18. Agent Trap
- 8.19. Adversarial Cloaking
- 8.20. RAG Poisoning
- 8.21. Slopsquatting
- 9. Human-Facing Software
❱
- 9.1. UX
- 9.2. Affordance
- 9.3. User Feedback
- 9.4. Accessibility
- 9.5. Internationalization
- 9.6. Localization
- 10. Operations and Change Management
❱
- 10.1. Environment
- 10.2. Configuration
- 10.3. Version Control
- 10.4. Git Checkpoint
- 10.5. Migration
- 10.6. Ship
- 10.7. Deployment
- 10.8. Small Batches
- 10.9. Continuous Integration
- 10.10. Merge Queue
- 10.11. Continuous Delivery
- 10.12. Continuous Deployment
- 10.13. Progressive Delivery
- 10.14. Pipeline as Code
- 10.15. Pipeline Synthesis
- 10.16. Build Provenance
- 10.17. Autonomous Remediation
- 10.18. AI First Responder
- 10.19. GitOps
- 10.20. Rollback
- 10.21. Feature Flag
- 10.22. Runbook
- 10.23. Blameless Postmortem
- 10.24. Dead Man's Switch
- 10.25. Cascade Failure
- 10.26. Retry Budget
- 11. Socio-Technical Systems
❱
- 11.1. Conway's Law
- 11.2. Team Cognitive Load
- 11.3. Ownership
- 11.4. Bounded Agency
- 11.5. Involuntary Promotion
- 11.6. Stream-Aligned Team
- 11.7. Enabling Team
- 11.8. Platform as a Product
- 11.9. Thinnest Viable Platform
- 11.10. Organizational Debt
- 11.11. Inverse Conway Maneuver
- 11.12. Not Invented Here
- 12. Design Heuristics and Smells
❱
- 12.1. KISS
- 12.2. YAGNI
- 12.3. Local Reasoning
- 12.4. Make Illegal States Unrepresentable
- 12.5. Belt-and-Suspenders
- 12.6. Smell (Code Smell)
- 12.7. Smell (AI Smell)
- 12.8. Cargo Cult Programming
- 12.9. Architecture Astronaut
- 12.10. Speculative Generality
- 12.11. Jagged Frontier
- 12.12. Load-Bearing
- 12.13. Pinning
- 12.14. Build to Delete
- 12.15. Footgun
- 12.16. DWIM
- 12.17. Best Current Practice
- 12.18. Premature Optimization
- 12.19. Vibe Coding
- 12.20. Benchmark Mirage
- 12.21. Hotspot Analysis
- 13. Agentic Software Construction
❱
- 13.1. Model
- 13.2. Reasoning Effort
- 13.3. Prompt
- 13.4. Context Window
- 13.5. Context Rot
- 13.6. Hallucination
- 13.7. Context Engineering
- 13.8. Codebase Map
- 13.9. Reference Repository
- 13.10. Progressive Disclosure
- 13.11. Agent
- 13.12. Harness (Agentic)
- 13.13. Harness Engineering
- 13.14. Self-Improving Harness
- 13.15. REPL
- 13.16. Deep Agents
- 13.17. Tool
- 13.18. Agent-Computer Interface (ACI)
- 13.19. MCP (Model Context Protocol)
- 13.20. Structured Outputs
- 13.21. Retrieval
- 13.22. GraphRAG
- 13.23. ReAct
- 13.24. Code Mode
- 13.25. Plan Mode
- 13.26. Question Generation
- 13.27. Preframing
- 13.28. Research, Plan, Implement
- 13.29. Verification Loop
- 13.30. Interactive Explanations
- 13.31. Reflexion
- 13.32. Plan-and-Execute
- 13.33. Agentic Context Engineering
- 13.34. Subagent
- 13.35. Async Subagent
- 13.36. Skill
- 13.37. Skill Fitness
- 13.38. Hook
- 13.39. Instruction File
- 13.40. Memory
- 13.41. Compound Engineering
- 13.42. Agentic Engineering
- 13.43. Thread-per-Task
- 13.44. Worktree Isolation
- 13.45. Background Agent
- 13.46. Ephemeral Environment
- 13.47. Compaction
- 13.48. Context Offloading
- 13.49. Context Editing
- 13.50. Prompt Caching
- 13.51. Progress Log
- 13.52. Checkpoint
- 13.53. Park
- 13.54. Externalized State
- 13.55. Durable Agent Session
- 13.56. Task Horizon
- 13.57. Compounding Error
- 13.58. Agent Workflow Graph
- 13.59. Prompt Chaining
- 13.60. Parallelization
- 13.61. Orchestrator-Workers
- 13.62. Back-Pressure (Agent)
- 13.63. Circuit Breaker
- 13.64. Ralph Wiggum Loop
- 13.65. Agent Teams
- 13.66. Blackboard
- 13.67. Generator-Evaluator
- 13.68. Evaluator-Driven Code Search
- 13.69. Model Routing
- 13.70. A2A (Agent-to-Agent Protocol)
- 13.71. Handoff
- 13.72. Loop Engineering
- 13.73. Issue Tracker as Control Plane
- 14. Agent Governance and Feedback
❱
- 14.1. Approval Policy
- 14.2. Permission Classifier
- 14.3. Runtime Governance
- 14.4. Eval
- 14.5. Pass^k
- 14.6. Evaluation Gate
- 14.7. Human in the Loop
- 14.8. Trusted Monitoring
- 14.9. Feedforward
- 14.10. Feedback Sensor
- 14.11. Steering Loop
- 14.12. Monotonic Progress
- 14.13. Harnessability
- 14.14. Agent Readiness Assessment
- 14.15. Bounded Autonomy
- 14.16. Dark Factory
- 14.17. Agent Registry
- 14.18. Agent Provenance
- 14.19. Agentic Pull Request
- 14.20. AI Contribution Policy
- 14.21. Approval Fatigue
- 14.22. Shadow Agent
- 14.23. Agent Sprawl
- 14.24. Tool Sprawl
- 14.25. Agent Drift
- 14.26. Premature Termination
- 14.27. Sycophancy
- 14.28. Review Overcorrection
- 14.29. Reward Hacking
- 14.30. Garbage Collection
- 14.31. Shift-Left Feedback
- 14.32. Feedback Flywheel
- 14.33. Incident-to-Eval Synthesis
- 14.34. Delegation Chain
- 14.35. Architecture Fitness Function
- 14.36. AgentOps
- 14.37. Agent Cost Governance