NubiferAI

AI-native cloud operations platform for NubiferOS

NubiferAI User Guide

NubiferAI is an AI-native cloud operations platform. Describe what you want to build, deploy, or change in natural language, and NubiferAI generates infrastructure-as-code, deployment scripts, and operational plans.

Getting Started — Installation, first launch, and your first nucleation
Providers — Setting up Claude, OpenAI, Ollama, and Bedrock
GUI Application — Walkthrough of the GTK desktop application
Seeds — AI personas for different cloud tasks
NubiferOS Integration — Using NubiferAI within NubiferOS

Quick Start

# Install (standalone)
pip install nubiferai

# Or via NubiferOS
nubi addon install nubiferai

# Launch the GUI
nubiferai-gtk

# Or use the CLI
nubiferai nucleate "Deploy a Python Flask app to AWS ECS with a load balancer"

How It Works

You describe what you want in natural language
A seed (AI persona) interprets your request with cloud expertise
An AI provider generates a structured plan called a nucleation
You review the actions, exports (Terraform, shell, CloudFormation), and safety checks
You execute (precipitate) when ready — or export for manual review

Key Concepts

Concept	Description
Nucleation	A structured plan of cloud actions generated by AI
Seed	An AI persona specialized for a cloud domain (AWS, Terraform, Kubernetes, etc.)
Precipitate	Execute a nucleation — run the generated infrastructure changes
NubiferTrail	Audit log of all nucleations and executions
Safety Classification	Controls what operations are allowed (sandbox/development/staging/production)

Interfaces

NubiferAI provides two interfaces:

GUI (nubiferai-gtk) — GTK4/libadwaita desktop application with visual seed browsing, provider configuration, and live execution output
CLI (nubiferai) — Terminal interface with Rich formatting, ideal for scripting and SSH sessions