
LifeOS — The AI-Powered Life Operating System
Repo: You do not have permission to view the full content of this post. Log in or register now.
Website: You do not have permission to view the full content of this post. Log in or register now.
Docs: You do not have permission to view the full content of this post. Log in or register now.
License: MIT
Stars: 18.2k+ | Forks: 2.4k+
What it is
A "general hill-climbing AI harness" built on top of Claude Code (and other agentic harnesses like Cursor/Codex) that gives your AI persistent memory, custom skills, and full context about your goals and life — so it works toward your "Ideal State" instead of starting from zero every session. Formerly known as PAI (Personal AI Infrastructure), renamed and rebuilt as a single self-contained skill.
Core ideas
- Current State → Ideal State — the whole system orbits this one loop, chasing what the author calls "Euphoric Surprise"
- Persistent memory ("Cortex") — hot-layer memory, a typed knowledge archive (people, companies, ideas, research), learnings, and work history that carries across sessions
- Skill system — one self-contained skill bundling a whole library: research, security, writing, art, and more
- The Algorithm — a 7-phase loop (Observe → Think → Plan → Build → Execute → Verify → Learn) with cross-vendor verification at higher complexity tiers
- Harness-agnostic — built on universal primitives (hooks, skills, context files), tested most on Claude Code but portable to any capable agent
- Security-hardened releases — deterministic gates (sanitizer-residue checks, license checks, dead-doc pointer checks) run before every public release
Install
Installed by an AI — just hand it the prompt:
Code:
Read https://ourlifeos.ai/install and install LifeOS for me.
Or, one-liner for Claude Code on macOS/Linux:
Code:
curl -fsSL https://ourlifeos.ai/install.sh | bash