LifeOS skill
Installer for Daniel Miessler's personal AI system: memory, hooks, a goals interview and a Pulse dashboard on top of Claude Code, Cursor or Codex
- 19,635 ★ on the repository
- Last commit
- MIT
Broad access
Why this level
- Installation writes hooks into the agent settings.json and places files under ~/.claude, and those hooks later run during your sessions
- The quick command pipes a script from the network into bash
- The system accumulates personal context: goals, contacts, notes and memory of past sessions
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/lifeosDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewBroad accessThe tool can change data, run commands or handle secrets.
- Installation writes hooks into the agent settings.json and places files under ~/.claude, and those hooks later run during your sessions
- The quick command pipes a script from the network into bash
- The system accumulates personal context: goals, contacts, notes and memory of past sessions
This is third-party code. Review the repository files before installing.
What it does
LifeOS is Daniel Miessler's set for turning a plain coding agent into an assistant that knows you and your goals. The repository ships as one skill, LifeOS, containing the installer, the setup, interview, update and uninstall workflows, and the whole file set that gets placed into the agent's configuration. Installation has two phases. First the agent detects your OS and harness, installs prerequisites, lays down templates and wires hooks, showing each change and waiting for your approval. Then it runs an interview: the assistant's name, your identity, your current state and your ideal state under the TELOS method. Your answers and any outside notes you provide go into the USER tree and the Pulse dashboard. Afterwards the system has persistent memory, routing of requests to the right skills and self-updates based on what it learns. The doctor command shows which capabilities work and which are broken, and proposes a fix.
Who it is for. For developers and founders who work in Claude Code, Cursor or Codex every day and want the agent to remember their goals, contacts and past decisions.
Good fit when
- The agent starts every session from zero and you are tired of re-explaining projects, goals and preferences
- You want one system on top of Claude Code with memory, hooks and a skill library rather than a pile of loose prompts
- You want to go through an interview about your current and ideal state and get a Pulse dashboard built from it
- You have notes and configs from earlier tools and want them pulled into the assistant context during installation
Not a fit when
- You need one narrow skill: LifeOS installs a whole system with hooks into the agent settings
- You work on a shared or managed machine where agent settings must stay untouched and bun cannot be installed
- You want a set of ready prompts for separate tasks: Fabric covers that
- Your agent has no hooks or skills support, which removes most of the point of the install
Example request
Read https://ourlifeos.ai/install and install LifeOS for me, asking before each settings changeLimitations
Needs bun and an agent on the level of Claude Code: the author tests the main path there and states compatibility with others through shared primitives. The one-line script targets macOS and Linux, other systems use the step-by-step install through the agent. User data lives in your agent configuration. The roadmap items in the README (local models, remote access, external notifications) are not built yet.
How to disable. The skill has an uninstall workflow: it shows what will be removed and deletes only LifeOS hook entries and system files by manifest. Your USER tree with identity, TELOS and memory stays, and removing it is a separate confirmed step. Before an upgrade the author advises backing up ~/.claude.
FAQ
How do I install LifeOS in Claude Code?
Paste "Read https://ourlifeos.ai/install and install LifeOS for me" to the agent. It reads the install page and walks the steps, asking permission before each change. For Claude Code on macOS and Linux there is a shortcut: curl -fsSL https://ourlifeos.ai/install.sh | bash. You need bun.
How is LifeOS different from Fabric?
Fabric is a collection of prompts for separate tasks and answers what to ask the AI. LifeOS is infrastructure around the assistant: memory, skills, routing, context about you and self-updates. The author describes the two as complementary, and many users plug Fabric patterns into their skills.
What is TELOS in LifeOS?
It is the description of your current and ideal state. The interview after installation collects it, and the system works to move you from the first to the second. A separate skill lets you edit TELOS after the first run.
What if the installation breaks something?
Copy ~/.claude before upgrading. The installer leaves the USER directory alone, merges settings rather than overwriting them, and detects an earlier install when you run it again. Git makes rollback easier.
Does LifeOS work only with Claude Code?
The author calls the project harness-agnostic: it is built on hooks, skills and context files. The install is described for Claude Code, Cursor, Codex, Hermes and OpenClaw, but the Claude Code path is the most tested.
Discussion
No questions or reviews yet.
Related
SkillsA starter skill kit for Claude Code and Codex: personal onboarding, a knowledge base, a system audit and a visual map of knowledge
Personal agent context in plain Markdown files you own, plus an onboarding skill built on a short interview
A mentor skill for Claude Code that guides you to LeetCode solutions with hints instead of handing over code at once
A library of 229 skills for work and life: sales, marketing, finance, engineering, design and parallel agent runs
A set of everyday agent skills: web search, browser automation, Gmail, Google Calendar and Drive, speech transcription and YouTube transcripts
Six profession bundles with 170 skills in total, each installed into Claude Code, Cursor or Codex with one command
Comparison with similar tools
Entries that share the main search query, and how they differ.
| Entry | Official | Repository stars | Main difference |
|---|---|---|---|
| LifeOS (this entry) | no | 19,635 | |
| Fabric | no | 44,485 | a collection of ready prompts run from the terminal one by one, without memory or hooks |
| Second Brain Skills | no | 837 | a skill set for knowledge work, with no installer or goals interview |
README in short
The description opens with the idea of moving from a current state to an ideal one and explains that LifeOS is installed by the AI itself from one prompt. It then lists the core components: TELOS, the Algorithm, the skill and hook systems, the Pulse dashboard, voice, learning and security. The FAQ compares the project with a bare agent and with Fabric and describes recovery after a failure. The end holds a roadmap, community links and the release history.
README badge
Original SKILL.md
--- name: LifeOS version: 1.5.43 description: Install and onboard a user into LifeOS — the Life Operating System (current state → ideal state via TELOS + the Algorithm). The agentic installer detects your OS + harness, wires hooks with permission, scaffolds your USER tree, pulls in sources you provide, and runs the TELOS / current→ideal interview that seeds your Pulse dashboard. USE WHEN install LifeOS, set up LifeOS, lifeos setup, lifeos-setup, lifeos interview, onboard me, run the interview, integrate LifeOS into my harness, update LifeOS, uninstall LifeOS, first-time setup, lifeos doctor, check my install, what capabilities are broken. NOT FOR building or cutting a LifeOS release (private release tooling), editing TELOS after onboarding (use Telos / Interview), or LifeOS system maintenance (use the private maintenance skill). disable-model-invocation: true argument-hint: "[setup|interview|doctor|update|uninstall]" allowed-tools: Read, Write, Edit, Bash, Glob, Grep --- # LifeOS The install + onboarding surface for **LifeOS** — the Life Operating System. One command takes a stranger on any harness from nothing to a working, personalized install whose Pulse dashboard already shows their current state vs ideal state — without making them adopt a whole new harness. ## How it ships LifeOS is distributed as **one self-contained skill** — the `LifeOS/` directory is the *entire* distribution. Everything ships inside it: the orchestrator (`SKILL.md`, `Workflows/`, `Tools/`), the whole-system payload under `install/`, and the one-line bootstrap at `install/install.sh`. **Nothing ships outside the skill** — no release-root `install.sh`, no `.claude/` clone. **The primary install is AI-native: give `INSTALL.md` (served at `ourlifeos.ai/install`) to your AI and say "install this."** LifeOS is AI-native, so the install is too — you hand the doc (or its link) to whatever harness you already use, and your AI installs LifeOS on your OS and harness, with permission at each step. It's the same document a human can read and follow. `INSTALL.md` opens with a capability gate, drives the install Tools (which run under `bun` on any OS, not a shell), wires integration per-harness (honest about what each gets), then runs Setup → Interview. A terminal shortcut stays for Claude Code on macOS/Linux: ``` curl -fsSL https://ourlifeos.ai/install.sh | bash ``` Both are served from the skill's own single sources of truth — `INSTALL.md` at the skill root, `install/install.sh` for the shell path (which hands off to the agentic `/LifeOS setup`). Two versions coexist and mean different things: the frontmatter `version:` is this skill's own **component** line (bumped by the maintainer-side `BumpSkillVersions`, which does not ship in the release), while the **distribution** version — what a user means by "LifeOS 7.x" — is the GitHub release tag and the `LIFEOS_RELEASES/<version>/` parent dir. Never read the component line as the release number. The payload (skills, hooks, system prompt, Algorithm, docs, runtime tools) rides along under `install/` and is placed during setup, with permission. ## Workflow Routing | Trigger | Target | |---------|--------| | `setup`, `/LifeOS setup`, "install LifeOS", "integrate into my harness" | `Workflows/Setup.md` | | `interview`, "onboard me", "run the interview", TELOS capture | `Workflows/Interview.md` | | `doctor`, "check my install", "what's broken", "what capabilities are live" | run `bun <configRoot>/LIFEOS/TOOLS/Doctor.ts` (see `INSTALL.md`) | | `update`, "update LifeOS", after a version bump | `Workflows/Update.md` | | `uninstall`, "remove LifeOS" | `Workflows/Uninstall.md` | `doctor` is the one tool-backed route — it needs no workflow because `Doctor.ts` is self-describing: it prints the four capability states (live / broken / declined / stale) and the exact fix command for anything broken. Relay its table, offer the fix it names, and honor `decline` — a declined capability is a legitimate way to run LifeOS, never a defect to nag about. Default flow (`/LifeOS setup`): **Setup phase** (system integration) → transitions into **Interview phase** (life onboarding). One continuous experience, two clearly-marked phases — setup is logistics, interview is meaning. Setup ALWAYS runs first; hooks must be wired before the interview seeds anything.