Available

The Three-Body Problem of Classified AI: Implementation Edition

Architecture and Verification

The Implementation Edition delivers the eight-property framework (P1-P8), the governance engine architecture, and the evidence pipeline that makes continuous authorization possible in classified environments.

Buy

Purchase links will appear here at launch.

Inside this volume

The P1-P8 Framework

Eight operational properties that define trust boundaries for AI agents in classified environments. Each property is verifiable, measurable, and maps directly to existing federal compliance frameworks.

Table of contents

What you'll read inside Volume II-B

01Chapter 8: Codebase Preconditions
02Chapter 9: Knowledge Base Integrity
03Chapter 10: Model Supply Chain
04Chapter 11: Governance of Noise
05Chapter 12: Drift and Decay
06Chapter 13: Design Principles for Operational Trust
07Chapter 14: Identity, Intent, and Chain of Custody
08Chapter 15: The Delegation Token and Activity Record
09Chapter 16: The Agentic RAG Architecture
10Chapter 17: The Governance Engine
11Chapter 18: Specialized Agents and Routing
12Chapter 19: Multi-Developer Agent Orchestration
13Chapter 20: Federated Trust Across Boundaries
14Chapter 21: Edge-Native Autonomy
15Chapter 22: The Harness Doctrine
16Chapter 23: P1-P8 as Harness Constraints
17Chapter 24: The Three Failure Modes
18Chapter 25: From Governance to Evidence
19Chapter 26: The Evidence Stream
20Chapter 27: Policy-as-Code and Risk Registers
21Chapter 28: The Authorization Decision Framework
22Chapter 29: The AO's Operating Picture
23Chapter 30: Trust But Verify

Access supplemental book materials

The Harness Doctrine/Volume II-B