Evolutionary Model

An AI agent that doesn't learn is just an expensive chatbot.

The Core Idea

Most people set up AI assistants once and use them forever the same way. The Evolutionary Model is different: the agent grows smarter with every session, accumulates skills, and becomes increasingly specific to its owner's needs.

The model has three axes of evolution:

CODEBLOCK0

---

Architecture

Layer 0 — Identity

Who the agent is. Fixed at birth, rarely changed.

CODEBLOCK1

Layer 1 — Memory

How the agent persists across sessions.

CODEBLOCK2

Key principle: no mental notes. If it's not written to a file, it doesn't exist after session restart.

Layer 2 — Skills

What the agent can do. Each skill is a self-contained capability module.

CODEBLOCK3

when_to_use is critical. Without it, the agent doesn't know when to activate the skill. Format:
CODEBLOCK4

Layer 3 — Protocols

How the agent behaves reliably. Learned from mistakes.

CODEBLOCK5

---

How Evolution Works

Session → Memory

Every session, the agent:

1. 1. Reads SESSION-STATE.md (hot context)
2. Reads today's daily log
3. Works
4. Writes new decisions/insights to daily log
5. Periodically distills into INLINECODE2

Task → Skill

When the agent solves a new type of problem:

1. 1. Documents the solution
2. Creates INLINECODE3
3. Adds when_to_use so it auto-activates next time

Mistake → Protocol

When the agent makes a mistake:

1. 1. Analyzes root cause
2. Adds rule to AGENTS.md or INLINECODE6
3. Future sessions inherit the fix

---

Skill Quality Standards

A skill is production-ready when it has:

• - [ ] when_to_use frontmatter — agent knows when to use it
• [ ] description frontmatter — discoverable in skill catalogs
• [ ] No hardcoded personal context (paths, names, tokens)
• [ ] config.json or env vars for user-specific settings
• [ ] README.md explaining what it does and how to configure
• [ ] Scripts that work from any machine (no absolute paths)

---

Starter Kit

Minimum viable agent setup:

CODEBLOCK6

Bootstrap checklist:

1. 1. Fill USER.md with owner's name, timezone, communication style
2. Write SOUL.md — personality takes 30 minutes, saves 1000 future corrections
3. Pick 3 starter skills from the catalog
4. Run first session — agent reads all files and introduces itself
5. After session: review what the agent wrote to memory files

---

The Compounding Effect

Month 1: agent knows your name and timezone
Month 2: agent knows your projects, communication style, key contacts
Month 3: agent anticipates needs, runs proactive checks, catches mistakes
Month 6: agent has accumulated skills specific to your workflow
Month 12: agent is irreplaceable — it carries institutional knowledge no new model can replicate

This is why the model is called "evolutionary": the value grows non-linearly. Not because the base model gets smarter, but because the accumulated context, skills, and protocols become a moat.

---

Why Not Just Use ChatGPT?

	ChatGPT / Standard Assistant	Evolutionary Model
Memory	Resets every session	Persists across sessions
Skills

Fixed capabilities | Grows with use |
| Context | Generic | Specific to you |
| Mistakes | Repeated | Documented + prevented |
| Value over time | Flat | Compounding |
| Portability | Locked to provider | Files you own |

The Evolutionary Model runs on any AI provider. The intelligence isn't in the model — it's in the accumulated files. You own them.

---

Contributing Skills

Skills are just markdown files. To share a skill:

1. 1. Remove all personal context (names, paths, tokens)
2. Replace with ${VARIABLE} or config.json entries
3. Add when_to_use frontmatter
4. Write a INLINECODE16
5. Submit to ClaWHub or share as a repo

---

See Also

• - SOUL.md — agent identity template
• INLINECODE18 — operating protocols
• INLINECODE19 — proactive check-in system
• Skills catalog: INLINECODE20