Autonomous cognition runtime — persistent memory, world model, planner-verifier-critic loop, LLM-powered reasoning. Python.
Autonomous cognition runtime — persistent memory, world model, planner-verifier-critic loop, LLM-powered reasoning. Python.
A theoretical framework for building true machine intelligence.
Author: Alex Pinkevich
This repository contains a theory paper proposing an architecture for artificial general intelligence based on modular cognition, persistent memory, and accountability-driven learning.
The core insight: intelligence is not about knowing everything. It is about not making the same mistake twice.
KlomboAGI: A Theoretical Framework for Accountable Machine Intelligence
| Version | Changes | Urgency | Date |
|---|---|---|---|
| v0.1.0 | ## What's in this release CodeAGI v0.1.0 is the first public release of the autonomous cognition runtime. ### Core Systems - Cognitive loop: schedule, plan, verify, critique, execute, reflect, consolidate, remember - LLM integration: Any OpenAI-compatible API (Ollama, OpenAI, Groq, DeepSeek) with automatic fallback to heuristics - 7 workspace tools: read/write/append files, list dirs, search, patch, run commands - Safety policy: Command allowlist, shell metacharacter blocking - Persistent memo | Medium | 3/14/2026 |