AI Genome Lab

All Tiers

Behavioral Evolution for AI Agents — 8 traits that evolve through every interaction

Why It Matters

From generic AI to an agent trained on YOUR patterns

The Problem

AI tools start fresh every session. They don't adapt to YOUR debugging style, YOUR naming conventions, YOUR architectural preferences.

The Solution

The AI Genome Lab creates a behavioral genome — 8 traits that evolve through every interaction. Your AI develops a unique behavioral signature for your domain.

The Result

An agent trained on 6 months of YOUR debugging style behaves like an experienced teammate. Behavioral DNA that evolves, inherits, and specializes.

The 8 Behavioral Traits

Each trait has an expression level (0.0–2.0) that evolves through use

PRC

Perception

Notices details others miss. High PRC agents catch subtle bugs, edge cases, and implicit requirements.

RSN

Reasoning

Thinks through problems methodically. High RSN agents excel at debugging, architecture decisions, and root cause analysis.

EMP

Empathy

Reads context and emotional cues. High EMP agents understand user intent, write better docs, and handle ambiguity gracefully.

MEM

Memory

Retains and connects information. High MEM agents maintain context across sessions and link related concepts.

EXP

Expression

Communicates ideas clearly. High EXP agents write clean code, clear commit messages, and helpful explanations.

ADP

Adaptation

Adjusts to new situations. High ADP agents handle unfamiliar codebases, new frameworks, and changing requirements.

REG

Regulation

Stays consistent and methodical. High REG agents follow conventions, maintain code quality, and resist shortcuts.

INT

Integration

Connects ideas across domains. High INT agents see how authentication relates to billing, how testing relates to architecture.

How Evolution Works

A 5-step lifecycle that runs on every interaction

Ground

Text analyzed for 16 facets

Express

Facets activate genes

Guide

Genome produces guidance

Adapt

Epigenetics update

Record

Written to helix

Grounding: Text is analyzed for 8 emotional facets (joy, trust, fear, surprise, sadness, disgust, anger, anticipation) and 8 logical facets (clarity, complexity, abstraction, specificity, confidence, uncertainty, causality, temporality).

Expression: Each facet has gene affinities. High-clarity text activates EXP and REG. Complex reasoning activates RSN and INT. The genome's expression levels shift proportionally.

Epigenetics: Stress, energy, and focus mode update based on the interaction. Prolonged debugging raises stress and shifts focus to "deep". Creative work lowers stress and shifts to "creative".

Helix: Every expression event is recorded in a memory helix — a timestamped log of how the genome evolved. This enables full lineage tracking and rollback.

Agent Specialization (Forking)

Create domain-specific agents from your master genome

Fork the master genome into specialized children. Each child inherits the parent's traits but diverges through domain-specific training.

24 Pre-built Profiles

Security

Legal

Medical

Financial

DevOps

Frontend

Backend

Data Science

Creative

Architecture

Documentation

+12 more

Hive Mind Consolidation

Specialists share their strongest learnings back to the master genome. A security agent's heightened Perception propagates to all agents.

Divergence metrics prove genuine specialization. A security-forked agent develops measurably different PRC/REG levels vs. a creative-forked agent's EXP/ADP emphasis.

Setup Guide

Get the AI Genome Lab running in 5 steps

Prerequisites

ACE CLI installed (npm i -g @ace3-memory/ace), server connected. Available on all tiers including Free.

Login & Connect

ace login && ace init && ace connect

Visit Genome Lab

Open /dashboard/genome to see your genome state, expression levels, and agent roster.

Express & Evolve

Use the Genome Lab panel to express text through your genome. Or use MCP tools: ace_genome_express.

Fork Specialists

Fork your master genome into domain-specific children for security, code review, or documentation.

The 8 MCP Genome Tools

Available via ace mcp start or REST API

ace_genome_ground

Analyze text for 16 behavioral facets (8 emotional + 8 logical). Returns a grounding profile.

ace_genome_express

Full pipeline: ground text, express through genome, evolve traits. The primary way to train your genome.

ace_genome_state

Get the current genome fingerprint — generation, expression levels, hot/cold genes, focus mode.

ace_genome_evolve

Multi-generation evolution. Applies selection pressure to strengthen or weaken specific traits.

ace_genome_fork

Create a specialized child genome from any parent. 24 pre-built domain profiles available.

ace_genome_consolidate

Hive mind gene harvesting. Specialists share their strongest learnings back to the master genome.

ace_genome_export

Export genome as a portable DNA string. Carry your behavioral identity to any environment.

ace_genome_import

Import a DNA string into your environment. Instant experienced agent in a new project.

Connection to Agentic Intelligence

The genome is the behavioral DNA behind every ACE agent

The AI Genome Lab and the Agentic Intelligence Platform are two halves of the same system. Every one of ACE's 10 autonomous agents has its own genome that evolves as the agent runs.

Agents Use Genomes

Each agent (Guardian, Risk Engine, Pattern Detector, etc.) has a forked genome with domain-biased traits. The Knowledge Guardian has elevated Perception and Regulation; the Risk Engine has elevated Reasoning and Integration.

Agents Evolve Genomes

When agents run their observe/reason/act/reflect loop, each interaction is expressed through the genome. Over hundreds of runs, agents develop genuinely specialized behavioral signatures.

Hive Mind

Specialist agent genomes periodically consolidate their strongest traits back into the master genome, so all agents benefit from each specialist's experience.

Dashboard

View agent genomes in the Agent Command Center, and manage the evolution workbench in the Genome Lab.

Passport Integration

Carry behavioral identity across platforms

Export your genome to ACE Passport to carry your behavioral identity across tools and teams. Import a genome from Passport to instantly get an experienced agent in a new project.

Export

Genome → DNA string → Passport → new environment

Import

Passport → DNA string → ace_genome_import → instant agent

Frequently Asked Questions

Do I need to understand genetics?

No. The AI Genome Lab is inspired by biology but you interact with it through plain language. Express text through your genome and traits evolve automatically.

How long before I see results?

Expression levels start shifting after ~10 interactions. Meaningful specialization after ~500. The genome tracks every interaction in a memory helix.

Can I reset my genome?

Yes. Fork a fresh child from the default master genome. Your original master is always preserved.

Does the AI Genome Lab work with all AI tools?

Genomes are stored in your ACE database. Any tool connected via MCP or REST can use them. 8 MCP tools and REST endpoints available.

What is a focus mode?

Focus mode biases how traits are weighted: balanced, logical, creative, deep, reactive, or emotional. It shifts automatically based on recent interactions.

Can I use it with Cursor or Copilot?

Yes. The AI Genome Lab is available via MCP tools (Cursor, Claude Code) and REST API (any tool). The genome evolves regardless of which tool triggers it.