✦ Aura OS

Terminal-Native AI Operating System — Project Z×Claw

Everything happens inside your terminal. No browser, no GUI, no pop-ups. Chat with AI, manage tasks, read email, browse the web, run agents — all from one place.

Install & Run

# Install globally
npm install -g aura-os

# Run
aura-os

# Or use npx (no install needed)
npx aura-os

From Source

git clone https://github.com/aniketkrs/aura-os.git
cd aura-os
npm install
npm start

On first launch, Aura OS runs onboarding: name, role, purpose, and PIN setup. Every subsequent launch requires your PIN.

✨ Features

• 🦕 Cute boot animation — dino greets you on startup
• 💬 Natural language — type "check my mail" instead of /mail inbox
• 🤖 AI Chat — OpenAI, Anthropic, Gemini, Mistral, or local Ollama
• 📧 Email — read inbox, compose, send (IMAP/SMTP)
• 🌐 Web Browser — browse any URL right in the terminal
• 📋 Tasks — create, track, and complete tasks
• 📅 Calendar — view upcoming events
• 🧠 Agents — 8 built-in agents + custom agent builder
• 🔒 Secure — PIN auth, encrypted storage, API key protection
• 🎨 Unique ASCII banners — each agent has its own art & color

Commands

| Command | Description | |---------|-------------| | /help | Full command reference | | /dash | Dashboard | | /chat [msg] | Chat with Aura LLM | | /task add <title> | Create a task | | /task list | List all tasks | | /mail inbox | View email inbox | | /mail send | Compose email | | /browse <url> | Browse URL in terminal | | /search <query> | Web search | | /agent start <name> | Start an agent | | /model <name> | Switch LLM model | | /status | System status | | /quit | Exit |

Natural Language (no slash needed!)

| Just type... | Does... | |---|---| | check my mail | Opens inbox | | show my tasks | Lists tasks | | search for react | Web search | | open youtube.com | Opens browser | | who am I | Shows profile |

🤖 Built-in Agents

| Agent | Description | |-------|-------------| | Memory Keeper | Consolidates and persists context | | Task Tracker | Monitors tasks, surfaces high-priority items | | Research Agent | Background research tasks | | Briefing Agent | Daily executive briefings | | Calendar Agent | Meeting reminders | | Writing Agent | Grammar and style suggestions | | Study Agent | Study session tracking |

LLM Models

Supports multiple providers — set API keys with /apikey <provider>:

| Provider | Models | |----------|--------| | Gemini | gemini-2.0-flash, gemini-2.5-pro | | OpenAI | gpt-4o, gpt-4o-mini | | Anthropic | claude-sonnet-4-5, claude-haiku | | Mistral | mistral-large, mistral-small | | Ollama | Any local model |

Security

• PIN required on every launch
• PBKDF2 hashed (100k rounds, SHA-512)
• 5 failed attempts → 5-minute lockout
• API keys stored in ~/.aura/.keys.json with chmod 600
• Session destroyed on exit
• API key leak detection in chat input

Roles

| Role | Auto-started agents | |------|---------------------| | Developer | memory-keeper, task-tracker | | Researcher | memory-keeper, research-agent | | Executive | memory-keeper, briefing-agent, calendar-agent | | Designer | memory-keeper | | Student | memory-keeper, study-agent, research-agent | | Writer | memory-keeper, writing-agent |

Requirements

• Node.js ≥ 18
• At least one LLM API key (or Ollama for local)

License

MIT © Aniket Kumar