Naming & AAEP Protocol

Why "Archon"

Archon (/ˈɑːrkɒn/) comes from ancient Greek ἄρχων, meaning "chief magistrate" — the single ruler of a city-state with clear authority boundaries.

Ancient Greece	Archon Protocol
One archon governs the polis	One agent governs the project
Laws constrain the polis	Constraint skills constrain the agent
Advisors provide info, not decisions	Internal skills provide analysis, not decisions
The archon is fully accountable	The agent owns the entire delivery

AAEP: AI Architect Evolution Protocol

AAEP defines how an AI agent evolves into the project's architect through continuous constraint-execution-evolution cycles.

The Gap AAEP Fills

AI coding tools provide capability (code generation, file operations, terminal execution). They don't provide methodology (workflow, quality assurance, experience accumulation).

AAEP is the methodology layer:

AI tools provide:  Capability
  ├── Code generation
  ├── File operations
  ├── Terminal execution
  └── Context understanding

AAEP provides:    Methodology
  ├── Constraint hierarchy
  ├── Delivery workflow
  ├── Evolution mechanism
  └── Governance model

AAEP's Four Layers

Layer	Components	Purpose
Constraint	Constraint skills	Hard boundaries (❌ prohibitions)
Workflow	Command agents/skills	Standard action sequences
Evolution	demand Stage 3.6, refactor	Self-strengthening mechanism
Knowledge	docs/, archon.config.yaml	Project memory

Evolutionary Pressure

Like biological evolution, AAEP creates mutation, selection, and heredity:

Biology	AAEP	Implementation
Mutation	Each /archon-demand produces new code	Stage 1
Selection	6-dim audit filters good and bad patterns	Stage 3
Heredity	Good patterns → rules, bad patterns → prohibitions	Stage 3.6

Key difference: biological evolution is random. AAEP evolution is directed — always toward fewer bugs, better performance, more consistent architecture.

Generation 1: Base constraints (file size, type safety)
    ↓ N tasks
Generation 2: Base + project-specific anti-patterns
    ↓ N tasks
Generation 3: Base + anti-patterns + performance + architecture
    ↓ N tasks
Generation N: Approaches the project's "perfect constraint set"