Summary

LumiOS is a personal AI operating system designed to incubate from user data, embodying their memories and personality. It aims to provide a persistent, cross-terminal AI companion that integrates across digital and physical spaces, offering a deeply personalized experience beyond typical AI assistants.

How It Works

The core features a personal AI that incubates from user chat logs and language samples, possessing persistent, evolving memory and a configurable personality engine. It supports 11 LLM providers (e.g., DeepSeek, OpenAI, Google) and integrates a comprehensive MCP ecosystem with 37 skills and 26 tools for diverse tasks. Cross-terminal synchronization ensures continuity across devices, with local offline operation. Interactions are visualized on an infinite canvas workbench, mapping user requests to tool calls and AI outputs.

Quick Start & Requirements

Installation involves cloning the repo, editing the .env file with API keys, and running platform-specific deployment scripts (Windows PowerShell, macOS/Linux bash). A minimal setup requires only DeepSeek API key and JWT secret, followed by npm install && npm run tauri:dev. Prerequisites include Node.js (18+), npm, Rust (for desktop builds), and Git. User data is stored in ~/LumiOS/.

Highlighted Details

• Personalized AI Incubation: AI personality and memory cultivate from user data for a unique, evolving digital extension.
• Extensive LLM & Modality Support: Integrates 11 major LLM providers and multiple TTS/STT engines.
• MCP Ecosystem & Skill Market: Features 37 MCP skills and 26 built-in tools, expandable via a community market.
• Cross-Terminal & Offline Operation: Seamless synchronization across devices with local offline functionality.
• Visualized Workflow: Canvas Workbench offers an infinite canvas for visualizing AI interactions and task flows.
• Desktop Automation: Utilizes visual models for screenshot-driven desktop control.

Maintenance & Community

The project seeks a CTO with AI engineering expertise. Community contributions via Issues and Pull Requests are welcomed. A roadmap shows significant feature completion, with mobile apps and advanced incubation planned. Founder contact details are provided.

Licensing & Compatibility

Licensed under GNU Affero General Public License v3.0 (AGPL-3.0). Modifications used in network services must be open-sourced under AGPL-3.0. Commercial closed-source licensing requires direct contact with the copyright holder.

Limitations & Caveats

Mobile applications and full holographic hardware integration are experimental or future developments. The advanced "passive perception" incubation mechanism is incomplete. AGPL-3.0 imposes strong copyleft obligations for network service usage.