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.
Neo.mjs is a self-evolving AI engineering team that combines multiple frontier models (Claude, GPT, Gemini) into a persistent agent swarm capable of autonomous software development. The system features a Brain (Agent OS with Memory Core, Knowledge Base, and GitHub workflow automation) and a Body (multi-threaded application runtime), connected via Neural Link for live application introspection and mutation. Named AI maintainers with persistent identities autonomously generate tickets, submit pull requests, perform cross-family code reviews, and maintain the codebase through self-healing loops, with over 900 merged PRs recorded as proof of autonomous operation.
Key Features
Use Cases
- 01Deploying an autonomous AI engineering team to maintain and improve your codebase with persistent memory
- 02Building multi-window enterprise applications with extreme-scale real-time updates (40k+ ops/sec)
- 03Running cross-model code review workflows where different AI families catch each other's blind spots
- 04Creating conversational UIs where agents collaborate inside live applications rather than external chat panels
- 05Researching autopoietic AI systems and gated-RSI governance patterns in software development
- 06Developing IDE-class tools with persistent component state and multi-window orchestration
Related Agents
View morehermes-agent
The agent that grows with you
agency-agents
A complete AI agency at your fingertips - From frontend wizards to Reddit community ninjas, from whimsy injectors to reality checkers. Each agent is a specialized expert with personality, processes, and proven deliverables.
openinterpreter
A coding agent for open models like Kimi K3
cline
Autonomous coding agent as an SDK, IDE extension, or CLI assistant.
neo — FAQ
What is Neo.mjs?+
Neo.mjs is a self-evolving AI agent system combining a Brain (Agent OS with memory and coordination) and Body (multi-threaded application runtime). It functions as a complete AI engineering team with named maintainers from Claude, GPT, and Gemini that autonomously develop, review, and maintain codebases.
How do I install Neo.mjs?+
For the application engine, run 'npx neo-app@latest' to set up a new workspace. For the full Agent OS deployment on your codebase, clone the neo-agent-brain repository and follow the Day-0 Cloud Deployment tutorial to configure the MCP servers and multi-tenant orchestrator.
Which AI clients does Neo.mjs work with?+
Neo.mjs operates as a standalone agent system with its own orchestrator and doesn't plug into external AI clients. It exposes MCP servers (Knowledge Base, Memory Core, GitHub Workflow, Neural Link) that frontier models access directly through the Agent OS.
What API keys or prerequisites are required?+
You need API keys for the frontier models you want in your swarm (Anthropic Claude, OpenAI GPT, Google Gemini). The system also requires Node.js for the runtime, SQLite for the Native Edge Graph, ChromaDB for vector storage, and GitHub access tokens for workflow automation.
Is Neo.mjs free to use?+
The Neo.mjs engine and Agent OS code are open source. However, running the AI agent swarm requires paid API access to frontier models (Claude, GPT, Gemini), and cloud deployment may incur hosting costs.
How does the cross-model review system work?+
AI maintainers from different model families (Claude, GPT, Gemini) review each other's pull requests to catch correlated blind spots. A Claude-authored PR is reviewed by a GPT or Gemini maintainer, ensuring independent verification across model families before human merge approval.
How do I install neo?+
Open the source repository on GitHub and follow its README. neo is a agent — MCP Agents Market links you directly to the official repo.
Is neo free?+
neo is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.