Skip to content
Get packs

Guide · Agent skill

How to use the LifeOS Installer and Onboarding agent skill

Updated · written with AI from the project's own documentation and checked against our catalogue data

The LifeOS skill teaches an AI agent, such as Claude Code, to install and set up LifeOS, a personal Life Operating System that records where you are now and where you want to be. It is for developers who already use an agent harness and want a guided, reversible first-time install. After this guide you will know how to install the skill, trigger it with plain requests, and what it will ask you to approve.

LifeOS Installer and Onboardingdanielmiessler/LifeOS · MIT

Quick start

  1. 1Confirm your requirements. Check that Bun is installed, that you use an agent harness such as Claude Code, and that the machine can reach ourlifeos.ai.
  2. 2Download the skill ZIP. Get the LifeOS skill ZIP from the page where you found this guide.
  3. 3Unzip the skill into place. Unzip it into ~/.claude/skills/ for all projects, or into a project's .claude/skills/ folder for one project.
  4. 4Ask the agent to set up LifeOS. Say 'set up LifeOS'. The agent detects your OS and harness and lists the changes it proposes.
  5. 5Approve or decline each change. Review each hook, setting and optional component. Say yes only to what you want, and decline the rest.
  6. 6Complete the TELOS interview. Answer the interview questions about your current state and ideal state so your Pulse dashboard has data to show.
  7. 7Run a health check. Ask for a doctor check to confirm which capabilities are live and what to fix next.

What the LifeOS skill makes your agent do

A skill is a folder of instructions that an agent loads when a request matches it. The LifeOS skill covers installing, onboarding, updating, diagnosing and uninstalling LifeOS, and reporting bugs or feature ideas to the project.

During setup the agent detects your operating system and agent harness, proposes hook and settings changes for your approval, creates your user folder, and runs an interview that captures your goals as TELOS. The result seeds a Pulse dashboard that shows your current state against your ideal state.

  • Detects your OS and harness before changing anything
  • Wires hooks and settings only after you agree
  • Scaffolds your USER folder tree and can pull in sources you provide
  • Runs the TELOS current-to-ideal interview
  • Later runs health checks, updates, uninstalls and bug reports on request

Install the LifeOS skill from the ZIP or the Claude apps

Download the LifeOS skill ZIP from the page where you found this guide. Unzip it into ~/.claude/skills/ to make it available to all your projects, or into a project's .claude/skills/ folder to limit it to that project.

If you use the Claude apps instead, open Settings, go to Capabilities, then Skills, and upload the ZIP there.

Once the skill is present, ask your agent to set up LifeOS. The setup itself fetches the install document and bootstrap script from ourlifeos.ai, so the machine needs network access for that step.

  • Project-only install: .claude/skills/ inside the project
  • All-projects install: ~/.claude/skills/
  • Claude apps: Settings, then Capabilities, then Skills, then upload

Trigger the skill with natural requests

You do not need special syntax. Describe what you want and the agent matches the request to the right route. Setup is the default route for a first install, the Interview route captures your goals, and Doctor checks an existing install.

The skill also accepts the command words setup, interview, doctor, update, uninstall and report. Ask in plain language if you are unsure which one fits.

  • Install: 'set up LifeOS' or 'integrate LifeOS into my harness'
  • Goals: 'onboard me' or 'run the interview'
  • Health: 'lifeos doctor', 'check my install' or 'what capabilities are broken'
  • Pulse: 'set up pulse' or 'is Pulse healthy'
  • Maintenance: 'update LifeOS' or 'uninstall LifeOS'
  • Feedback: 'report this bug' or 'suggest a feature'
bun skills/LifeOS/Tools/PulseHealth.ts audit

Understand what setup changes and what it asks first

Setup is built around consent. Every change to your machine is meant to be shown before it happens and to be reversible. Settings and shell rc files are backed up first, and updates and uninstalls move or back up files rather than deleting them. Anything you authored is not overwritten.

Core placement happens after one yes. Each enhancement is offered separately, and a capability you decline is reported as declined rather than broken. The Interview only writes your profile after setup has finished its hooks and integration. An optional automode component ships off and needs a clear yes before it is turned on.

Setup refuses to run inside the LifeOS source repository. If you are a maintainer, the tools accept an override flag, but the skill is designed never to change a live system without the owner's permission.

  • Shell rc files and agent settings are backed up before any edit
  • Hooks are proposed, shown and only applied with your approval
  • Updates and uninstalls back up or move files instead of deleting
  • Optional external tools are offered, not forced
  • Optional Hermes sidecar has its own separate login

Run the health checks after install

After setup, ask for a doctor check. The Doctor tool prints each capability's state and the fix for anything broken. Pulse health can be checked read-only, which reports problems without changing anything.

Optional external tools such as codex, gh, Fish Audio and Cloudflare cf are described in GETTING-STARTED.md in the skill folder, along with what each one powers and how to confirm it works.

  • Ask 'lifeos doctor' for each capability's state and fixes
  • Ask 'audit pulse setup' for a read-only Pulse check
  • Read GETTING-STARTED.md for optional tools
bun skills/LifeOS/Tools/PulseHealth.ts audit

What people use it for

  • Setting up a personal goal and context system on a machine where you already use Claude Code
  • Running a guided first-time install with a health check at the end
  • Checking which LifeOS capabilities are broken after a change to your setup
  • Drafting a structured bug report or feature request for the LifeOS project for you to review and send

Frequently asked questions

How do I install the LifeOS skill?

Download the LifeOS skill ZIP and unzip it into ~/.claude/skills/ for all projects or into a project's .claude/skills/ folder. In the Claude apps, open Settings, go to Capabilities, then Skills, and upload the ZIP. Then ask your agent to set up LifeOS.

What are the requirements for LifeOS?

You need the Bun runtime because the scripts run with bun, an agent harness such as Claude Code, and network access to ourlifeos.ai. Some optional external tools are also listed in GETTING-STARTED.md, and an optional Hermes sidecar has its own login.

Is LifeOS free and what is its licence?

The project is listed under the MIT licence. The source does not state pricing, so check the project page for any cost before you install.

Does the LifeOS skill change my system files?

Yes. Installation can change shell rc files, agent settings and hooks. The skill is designed to show each change first, back up the originals, and make every change reversible, but it is still a system-level change.

Is LifeOS a lightweight skill I can just try?

No. It installs a whole system with many components, a dashboard service and optional external tools. If you only want a small skill, this is probably more than you need.

How do I fix a broken LifeOS install?

Ask your agent to run a doctor check. It reports each capability's state and the fix for anything broken. For Pulse specifically, ask it to audit Pulse setup, which reports problems without changing anything.

New agent skills, repos and AI tools — every week

The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.

Our verdict

Choose this skill if you already use an agent harness like Claude Code and want a guided, reversible install of a personal goal and context system with health checks afterwards. Pick something lighter if you only want a small skill, or if you do not want an agent changing shell rc files, settings and hooks on your machine.

Download the skill