I’d love to share a project I’ve been building: AxiomOS v19.2 — a decentralized swarm of AI coding agents. Each agent has a specialization (e.g., speed, readability, robustness) and generates Python scripts in response to a task.
The twist: agents also gossip about their own reasoning, receive reputation scores, and evolve over generations based on fitness and ethical feedback.
A meta-agent called the AIOverseer synthesizes the best outputs and mutates the population’s DNA.
I’d love to share a project I’ve been building: AxiomOS v19.2 — a decentralized swarm of AI coding agents. Each agent has a specialization (e.g., speed, readability, robustness) and generates Python scripts in response to a task.
The twist: agents also gossip about their own reasoning, receive reputation scores, and evolve over generations based on fitness and ethical feedback.
A meta-agent called the AIOverseer synthesizes the best outputs and mutates the population’s DNA.
Core modules: – agent specialization (focus traits) – self-correcting LLM pipeline – gossip + reputation dynamics – guided synthesis & evaluation – Tkinter-based interface
This is not just a code generator — it’s an experiment in emergent ethics and algorithmic reflection.
Repo: https://github.com/Linutesto/AxiomOS Feedback, questions and ideas for collaboration welcome!