By ludwigawra
Your second brain for Claude Code — persistent memory, learned patterns, and skills that improve over time.
Show the user every skill AI-OS has installed, grouped by purpose, with one-line descriptions and when to use each. Use this any time the user asks "what can this do", "what commands are there", or "help" — and at the end of /aios-start to orient a fresh install.
The interactive guide to AI-OS. Three modes — tour (linear walkthrough of how the system works), status (live setup-progress with next-step suggestions), and ask (free-form Q&A about the system). Use this whenever the user asks "what is this", "how does this work", "how do I extend", "how do I add", or any orientation question. Also use when the user is stuck mid-setup.
Bootstrap a new AI-OS in the current directory. Conversational onboarding that scaffolds a brain-inspired personal operating system — persistent memory, learned patterns, and skills that improve over time. Use this the first time a user sets up AI-OS in a repo.
Non-destructive update for an installed AI-OS. Pulls new skills, hooks, and templates from the current plugin version into the user's `.claude/` directory without ever touching the brain folders (memory, learning, MEMORY.md, etc.). Use after the user upgrades the plugin (`npm update @aios/plugin` or plugin marketplace auto-update) to sync their local AI-OS with the new plugin version. Handles added skills, changed core skills, removed skills, new hooks, and template changes. Prompts before overwriting any file the user has modified.
Auto-generate intent-wrapped skills from a connected MCP's tool definitions. The user connects a new MCP (Gmail, Notion, Linear, Supabase, whatever), then runs /forge-skill <mcp-name> and the brain proposes 3-5 skill candidates that wrap those tools with natural-language intents. Every skill is approved by the user before it's written. This is how AI-OS grows capabilities as the user connects new tools.
Matches all tools
Hooks run on every tool call, not just specific ones
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Your second brain for Claude Code. Persistent memory, learned patterns, skills that improve over time.
AI-OS is a Claude Code plugin that installs a brain-inspired operating system on your machine. It gives Claude a persistent place to remember, reason, and improve — across every session, in every repo, for as long as you keep using it.
Install once. Keep everything.
Claude Code is powerful inside a session and forgetful between them.
AI-OS fixes this by giving Claude a file-system-native brain. A folder on your machine, modeled on the regions of a real brain, that Claude reads at the start of every session and writes to during it. Your knowledge stays local, versionable, and yours.
One folder, modeled on the brain but named in plain language. Each folder has one purpose. Each file has a place.
| Folder | What lives there |
|---|---|
CLAUDE.md | Identity, decision rules, executive control |
MEMORY.md | Long-term memory — user-triggered writes only |
risks.md | Risk flags — read before important actions |
memory/ | Decisions, sessions, transcripts |
learning/ | Corrections, patterns, skill feedback |
knowledge/ | People, companies, market notes |
projects/ | Active projects, each with its own CLAUDE.md |
routines/ | Daily, weekly, and on-demand routines |
blueprints/ | Templates and reusable how-tos |
voice/ | Brand and communication style |
system/ | Goals, architecture, vital signs, setup progress |
archive/ | Inactive but searchable |
No database, no cloud, no vendor lock-in. The folder is the brain — but you can read every file with a normal editor.
A curated set of skills that ship with the plugin and know how to use the brain.
memory-search — search the brain before starting workdecision-check — prevent contradictions by surfacing prior decisionsnightly-consolidation — process short-term memory, extract patterns, improve skillsmorning-briefing — daily prioritized summaryemail-triage — categorize and draft repliesmeeting-prep — surface relationship and decision contextrelationship-check — last contact, open commitments, next actionproject-status — progress, blockers, next actionsbrain-dump-content — turn raw thoughts into structured contentcontent-interview — Claude asks, you answer, drafts get madereflect — goal-by-goal progress and course correctionforesight — forward-looking prioritiessignal-calibration — the brain tuning its own signal detectoraios-start — one-shot install and setupaios-help — interactive guide: tour the system, check setup progress, ask anythingSkills improve over time. When you give feedback, the nightly consolidation edits the skill itself.
AI-OS is a Claude Code plugin distributed directly from this GitHub repo. No Anthropic marketplace submission, no gatekeeper, no shell script. Anyone with Claude Code installed can add it as a plugin marketplace in one line.
| Requirement | Check | Install |
|---|---|---|
| Claude Code | claude --version | docs.claude.com/en/docs/claude-code |
That's it. No git, no bash, no node, no python3 required.
In any Claude Code session, run:
/plugin marketplace add ludwigawra/AIOS
This tells Claude Code to treat this GitHub repo as a plugin marketplace. Claude Code clones it locally (cached), reads .claude-plugin/marketplace.json, and registers AI-OS as an installable plugin.
/plugin install aios@aios
(Format: plugin-name@marketplace-name — both happen to be aios here.) Claude Code installs the plugin's skills, hooks, and commands into your Claude Code config. Nothing touches your system outside of Claude Code's plugin directory.
Open a Claude Code session in the folder where you want your brain to live (examples use ~/my-brain), then run:
/aios-start
This is a conversational onboarding skill — Claude walks you through a short setup (name, role, company, which integrations you use), then scaffolds your brain folder: nine regions, templates, an initial MEMORY.md, a CLAUDE.md personalized to you, a live system/setup-progress.md checklist, and session hooks wired up.
npx claudepluginhub ludwigawra/aios --plugin aiosCore skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
Harness-native ECC operator layer - 67 agents, 271 skills, 92 legacy command shims, reusable hooks, rules, selective install profiles, and production-ready workflows for Claude Code, Codex, OpenCode, Cursor, and related agent harnesses
Superpowers Plus core skills library for Claude Code: planning, execution routing, TDD, debugging, and collaboration workflows
Claude harness - A harness for solo developers (Vibecoders) to handle full-cycle contract development.
Intelligent prompt optimization: injects the right context at the right moment so Claude lands a better first output. Clarifies vague prompts with research-based questions, plus targeted nudges for approach selection, plan readability, workflow routing, background execution, subagent routing, output readability, user-decision questions, and plan-mode assessment
Plugin-safe Claude Code distribution of Antigravity Awesome Skills with 1,561 supported skills.