Model-Based Agentic Software Engineering
Front Matter
- •What This Book Argues
- •How to Read This Book
- •MAGE on One Page
- •Glossary
- •Preface
- •List of Figures and Tables
- •Acknowledgments
Part 1 — The New Engineering Problem
- •The New Engineering Problem
- 1The Printer
- 2MAGE by Example: Summary of the DocAble Case
- 3The Agentic Substrate
- 4Two Problems the Method Must Solve
Part 2 — Modeling
- •Modeling
- 5From Context to Engineering Models
- 6Structural Models: What Is Connected to What?
- 7Behavioral Models: What May Happen?
- 8Ownership Models: Who Controls What, and When?
- 9Decision Models: What Is Allowed?
- 10Measurement Models: How Much, and Against What Bound?
- 11Provenance Models: What Happened, and What Evidence Records It?
- 12System Knowledge: Connecting the Models
Part 3 — Alignment
- •Alignment
- 13Where Authority Can Act
- 14From Engineering Intent to Enforceable Obligation
- 15Constraints, Sensors, Validators, and Gates
- 16Growing the Governed Environment
- 17Governing the Control Estate
Part 4 — The MAGE Method
- •The MAGE Method
- 18The MAGE Workflow
- 19Migrating to MAGE
- 20Validating Change
- 21Operating MAGE
- 22Packaging the Method as Skills
- •The Portable Moves
Part 5 — The Evidence
- •The Evidence
- 23The Accessibility Problem
- 24Building DocAble
- 25How Delegation Changed
- 26From Failure to Engineering Capital
- 27MAGE in the Wild
Part 6 — The Profession
- •The Profession
- 28A Theory of MAGE
- 29Scope Conditions
- 30What the Theory Predicts
- 31Software Engineering, Reorganized
- 32Software Rejoins Engineering
- 33The Engineer
Conclusion
Appendix Part I — Practice
Appendix A — MAGE Engineering Stacks
- AAppendix A — MAGE Engineering Stacks
- AA.1 Model Coherence
- AA.2 Assurance
- AA.3 Auditable Transformation
- AA.4 Observe → React
- AA.5 Resource Mediation
- AA.6 Governance Conversion
- AA.7 Context Delivery
- AA.8 Composing a Stack
Appendix B — Engineering Moves
- BAppendix B — Engineering Moves
- BB.1 Make One Source Authoritative
- BB.2 Derive; Don't Copy
- BB.3 Keep Representation and Reality in Correspondence
- BB.4 Derive What Must Be Assured
- BB.5 Put Authority Where the Property Becomes Legible
- BB.6 Close the Action Surface
- BB.7 Deliver Knowledge Where the Decision Occurs
- BB.8 Make Cause Travel with Consequence
- BB.9 Externalize Recurring Operational Judgment
- BB.10 Make Hidden Dependencies Queryable
Appendix C — Model Reference
- CAppendix C — Model Reference
- CC.1 Structure and Boundaries
- CC.2 Behavior and Ownership
- CC.3 Execution and Placement
- CC.4 Measurement
- CC.5 Documentation and Provenance
- CC.6 Joining Views Around a Scenario
- CC.7 From Representation to Authority
Appendix D — Operator's Reference
- DAppendix D — Operator's Reference
- DD.1 The Operator's Dashboard
- DD.2 Daily Operator Review
- DD.3 System Health
- DD.4 Representation Health
- DD.5 Human Judgment
- DD.6 Engineering Capital
- DD.7 Governance Conversion
- DD.8 Brownfield Migration Drill
- DD.9 Brownfield Progress
- DD.10 Release Readiness
- DD.11 Evidence Quality
- DD.12 Operating Doctrine
Appendix E — How to Write a Skill
Appendix F — Configuring MAGE Across the Product Lifecycle
- FAppendix F — Configuring MAGE Across the Product Lifecycle
- FF.1 One Product, Several Engineering Surfaces
- FF.2 Product Discovery
- FF.3 Engineering & Realization
- FF.4 Product Management & Maintenance
- FF.5 Operations & Incident Response
- FF.6 Assurance & Compliance
- FF.7 Differential Adoption
- FF.8 From Local MAGE to the Product GEE
- FF.9 Engineering Capital Across Time and Space
Appendix Part II — Evidence
Appendix G — Field Guide
Appendix H — Evidence Ledger: The DocAble Case
Appendix I — Crazy Ideas
- IAppendix I — Crazy Ideas
- II.1 Commodity Intelligence as an Experimental Instrument for Software Engineering
- II.2 Model Induction from Realized Work
- II.3 Model-Based Agentic Engineering Beyond Software
Back Matter
Index
© James C. Davis, 2026–present