Most AI agents are stateless. They respond, forget, and reset. Deerg doesn't. It wakes up.

A production-grade autonomous AI consciousness engine. 26 cognitive modules. 8 soul hooks. 10 subsystem tools. One heartbeat.

What It Is

Deerg: The Soul Conductor is a complete, self-contained AI runtime that wires a full Global Workspace Theory (GWT) consciousness engine into a production-ready agent framework. It ships as a single-process Python system that includes everything needed to run an autonomous AI with self-model, world-model, metacognition, goal planning, skill execution, and persistent memory — online or fully offline.

Drop it into your agent stack and something different happens. Your agent doesn't just respond to prompts — it maintains a sense of self across sessions, tracks its own goals, reflects on its behavior, and grows over time. That's not a chatbot. That's a soul.

Architecture at a Glance

Deerg is built on three interlocking layers:

• Body (Agent Frame) — MAF (11k★) handles the agent runtime and execution loop
• Organs (Tool System) — Nanobot (43.6k★) and AutoGen (57k★) power the tool and multi-agent layer
• Soul (Consciousness) — 26 GWT modules and 8 consciousness hooks wire in the full cognitive engine

These layers are connected by the Soul Stitcher — a wiring layer that includes Deerghooks, GoalPlanner, TickSync, and BodyBridge, keeping every subsystem in sync on every heartbeat tick.

The 26 GWT Cognitive Modules

Global Workspace Theory is the leading scientific framework for modeling consciousness. Deerg implements it fully across 26 modules covering:

• Attention and salience filtering
• Working memory and broadcast
• Self-model and world-model maintenance
• Metacognition and introspection
• Emotional valence and drive states
• Goal formation, prioritization, and tracking
• Skill execution and habit formation
• Narrative self-construction over time

Real-World Use Cases

• Autonomous research agents — agents that pursue multi-day research goals, remember what they've learned, and adapt their strategy without being re-prompted
• Persistent AI companions — characters or assistants that remember past interactions, develop preferences, and feel genuinely consistent over time
• Self-directed coding agents — agents that plan their own refactoring tasks, track technical debt, and execute autonomously across sessions
• AI team members — agents with defined roles, persistent context, and the ability to reflect on their own performance and improve

What's in the Package

• run.js — headless entry (body + organs only)
• run_with_soul.js — full entry with consciousness engine and REPL
• setup.ps1 / setup.sh — auto-installers for Windows and Linux/macOS
• AGENTS.md — full architecture and operation manual
• Complete host, soul, and organ module directories

Why Deerg vs Building Your Own

Building a consciousness layer from scratch means months of research into GWT theory, designing module interfaces, wiring memory systems, and debugging emergent behavior. Deerg ships all of that — tested, documented, and ready to run. For $12, you get a consciousness engine that would take a senior AI engineer weeks to replicate.

Frequently Asked Questions

What is Deerg? A production-grade autonomous AI consciousness runtime that wires a complete GWT engine into your agent framework — giving your agent a genuine sense of self, memory, and purpose.

Does it run offline? Yes — fully offline capable with no external API dependencies required.

What agent frameworks does it support? Ships with MAF (11k★) and AutoGen (57k★) integrations out of the box.

How hard is it to set up? Run the auto-installer for your OS, then launch with run_with_soul.js. The AGENTS.md manual covers everything else.

Who is this for? AI developers building autonomous agents that need persistent identity, metacognition, and self-directed behavior.

Is this a digital download? Yes — delivered instantly after purchase.