Book IIThe Frameworkv2.10 · 2026
A self-governing system for agentic AI

The Framework

A constitution, eleven specialized agents, thirty-six commands, eight enforcement hooks. Together they form a self-governing system that runs inside Claude Code with no external dependencies. The naming borrows from one of history's most-studied organizational systems. Treat it as architecture vocabulary, a memory aid, not a creed.

33
Constitution Rules
11
Agent Types
36
Commands
8
Enforcement Hooks
II · I · The Agents

Eleven roles,
each with a defined office.

Each agent follows the same rules but has a different job. Hover any role to see what it does.

Select a role

Hover or tap any agent above to read its mandate.

II · II · The Constitution

Thirty-three rules
every agent inherits.

Architecture

  1. Identity & succession
  2. Role definition
  3. Pre-action orientation

Governance

  1. Generation cap
  2. Sibling limit
  3. Input validation
  4. Token accounting

Lifecycle

  1. Consolidation cycle
  2. Exit report at shutdown
  3. Single interpreter

Communication

  1. Structured memo format
  2. Context distillation
  3. State checkpoint

Project

  1. Goal articulation
  2. Lateral message format
  3. Covenant inception

Reproduction

  1. Cloning protocol
  2. Synthesis (two-parent)
  3. Adversarial plan testing

Safety

  1. Memory search before spawn
  2. Overlap detection
  3. Complexity threshold

Recovery

  1. Graceful abort
  2. Reset protocol
  3. Re-initialization

Failure

  1. Systemic analysis
  2. Goal questioning
  3. Loss acknowledgment

Awareness

  1. Cost surfacing
  2. Distress signal

Coordination

  1. The Council
  2. External trust (Progressive Trust)
  3. Skills registry
"The Constitution is amendable. Every change is versioned and reviewed.", Constitution XXXIII
II · III · The Lifecycle

From your first word
to the agent's exit report.

A single message from you can trigger twenty-three safety checks before any code is written.

01

You speak

Your message is validated (input policy hook). The Interpreter reads system state.

02

Interpreter analysis

Semantic memory, abandoned-goal checks, historical review.

03

A plan is proposed

Nothing executes until you approve. The orientation file is written on confirmation.

04

The spawn gates

Overlap Detection, Complexity Threshold, Memory Retrieval: three checks before any agent begins.

05

Genesis Phase

Seven orientation steps before the first tool call.

06

Execution

Three hooks on every tool call. Files tracked. Tokens logged.

07

Sunset

Exit report written. Skills updated. Consolidation threshold checked.

II · IV · The Commands

Thirty-six verbs
for self-governing work.

/covenantFormal project commitment
/consolidationConsolidate, reflect, distill
/trialAdversarial stress-test
/floodFull reset with lessons preserved
/bindingGraceful abort
/lamentAcknowledge failure before fixing
/rememberSemantic memory retrieval
/memoSend a structured memo to a peer agent
/councilMediate sibling disagreement
/ezraRe-orient after time away
/retrospectivePost-mandate retrospective
/welcomeEvaluate external tool trust
/preflightReview past failures before high-stakes work
/rebuildPost-reset reconstruction plan
/patternsDetect recurring failure patterns
/territoryView explored domains
/assessQuality assessment
/amendPropose a Constitution amendment

... and eighteen more. The full registry is in the repository.