The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond.
Nuwax
Nuwax Agent OS - An enterprise-grade AI Agent Development and Operation Platform - Providing a complete solution for agent creation and distribution, knowledge base management,…
Install
docker run hello-worldclaude mcp add nuwax -- docker run hello-world{
"mcpServers": {
"nuwax": {
"command": "docker",
"args": [
"run",
"hello-world"
]
}
}
}Install command taken from the project's README or registry entry. Always review code before running it.
README excerpt
Read full README on GitHub →Nuwax
Nuwax AI - Easily build and deploy your private Agentic AI solutions.
Official Website: https://nuwax.com
Demo: https://agent.nuwax.com
中文文档|English Doc | Contributing | Documentation
Installation & Deployment
Use the official nuwax-cli command tool to quickly deploy services locally.
Quick Start
Environment Preparation
System Requirements
- System Requirements: Ubuntu 22.04 LTS or later (other Linux versions not fully tested), macOS 10.15+, Windows 10/11 (support coming soon)
- Hardware Requirements: 4 cores 8GB RAM or higher
- Environment Requirements: docker, docker-compose V2 environment Docker Installation Guide
Supported Platforms
- Linux: x86_64, ARM64
- Ubuntu 22.04 LTS (recommended)
- Current user needs Docker permissions, verify with
docker ps. If you encounter permission issues, you can run with sudo privileges. - Alibaba Cloud mirror acceleration is recommended
- macOS: Intel, Apple Silicon (M1/M2), Universal
- macOS 10.15 (Catalina) and later versions
- OrbStack is recommended (free for personal use, better performance)
- Ensure OrbStack or Docker Desktop is started
- First-time use may require allowing unknown developers: System Preferences → Security & Privacy
Local Deployment Service Guide
There are 2 deployment services:
- Main project service (required)
- Agent Computer (Sandbox) deployment guide (optional)
In the main local service, configure one or more Agent Computer deployment addresses to use the Agent Computer (Sandbox). Since the "Agent Computer (Sandbox)" includes a personal computer (sandbox) which requires more resources, it supports separate deployment across multiple servers.
Deploy Main Service
Agent Computer (Sandbox) Deployment Guide
You can deploy Agent Computer services on multiple different servers, achieving distributed agent sandbox capabilities through configuration.
Environment Requirements: Each server needs Docker and Docker Compose environment installed, refer to Docker Installation Documentation.
Common Management Commands
Service Management
- Start service:
./nuwax-cli docker-service start - Stop service:
./nuwax-cli docker-service stop - Restart service:
./nuwax-cli docker-service restart - Check status:
./nuwax-cli docker-service status
Backup Management
Backup service requires stopping Docker application servers, it's recommended to operate during business low-peak periods
- One-click Backup (Recommended):
- Manual backup execution:
./nuwax-cli auto-backup run - List all backups:
./nuwax-cli list-backups - Restore from backup:
./nuwax-cli rollback [BACKUP_ID]
- Manual backup execution:
Upgrade Management
Application service upgrade, using command ./nuwax-cli auto-upgrade-deploy run will automatically detect and download new versions for deployment.
Complete upgrade process:
Similar plugs
Reverse engineer anything with agents, from app behavior down to native binaries.
Build Agentic workflows, RAG pipelines, with rich AI model and tool support on one collaborative workspace. Deploy on cloud, VPC, or self-hosted, so teams move from prototype…
User-friendly AI Interface (Supports Ollama, OpenAI API, ...)
A cross-platform desktop All-in-One assistant for Claude Code, Codex, OpenCode, OpenClaw, Grok Build & Hermes Agent. Only official website: ccswitch.io