neomjs/neo
Neo.mjs is a self-evolving software organism: a professional end-to-end AI engineering team whose cross-model swarm inhabits live apps via Neural Link, Active Hybrid GraphRAG, DreamService, and self-healing loops.
What it solves
Neo.mjs is designed to move beyond stateless AI copilots by creating a "self-evolving software organism." It solves the problem of AI agents forgetting context between sessions and the lack of coordination between different AI models. It provides a framework where a swarm of AI agents from different labs (Claude, Gemini, GPT) can collaborate, review each other's work to eliminate blind spots, and inhabit a live application runtime to mutate state and code in real-time.
How it works
The project is split into two "hemispheres":
- The Brain (Agent OS): An AI-native toolchain that manages persistent memory via a Native Edge Graph and a Memory Core. It uses a
DreamServiceto distill noisy session data into a "Golden Path" topology. It coordinates a structured institution of named AI maintainers who handle issues, PRs, and reviews autonomously. - The Body (Application Engine): A high-performance, multi-threaded runtime (using App, VDom, Data, and Canvas Workers) that allows AI agents to "possess" the application. Through the Neural Link, agents can inspect the runtime state and hot-patch methods without browser reloads.
Who it’s for
- Engineers building high-scale, multi-window enterprise applications (e.g., financial trading desks) that require extreme performance (40k+ ops/sec).
- AI Architects seeking a substrate for multi-agent systems with persistent memory and the ability to mutate live runtimes.
- Researchers studying autopoietic systems and multi-agent organism governance.
Highlights
- Cross-Model Swarm: Uses agents from rival labs to perform peer reviews, reducing correlated errors.
- Neural Link: A possession interface that lets AI agents inhabit and mutate live application state.
- Agent OS: A persistent engineering institution with its own memory, identities, and autonomous GitHub workflow.
- MX (Model Experience): An evolution mechanism where internal agent friction is converted into tickets and skills to improve the system.
- Off-Main-Thread Architecture: A production-ready engine that prevents UI freezes by isolating logic, VDOM, and data processing into separate workers.