A free, role-agnostic AI Operating System template for organisations on the Microsoft stack. Fork it, set three org-level values once, and hand it to your team — each person runs
/onboardto make it theirs. The generic upstream isgithub.com/nateherkai/AIS-OS; this is the Microsoft variant.
This template ships blank. Three org-level placeholders need your values before rollout (do this
once in your fork, or let the first /onboard fill them):
| Placeholder | Replace with | Example |
|---|---|---|
{{ORG_NAME}} |
Your organisation's name | Acme Corporation |
{{ORG_DOMAIN}} |
Your work tenant domain (the account-check enforces this) | acme.com |
{{SHAREPOINT_HOST}} |
Your SharePoint host | acme.sharepoint.com |
⚠️ Keep your fork private if you fill these with real values. Your tenant domain and SharePoint host are internal info — don't publish a filled-in copy publicly. This upstream is generic on purpose.
A blank, role-agnostic AIOS template re-pointed for a Microsoft / corporate context. A typical
rollout starts with IT, then Supply Chain, then anyone. Each person personalises it via /onboard:
- MCP-preferred, CLI/REST fallback. Each connection uses its Microsoft MCP server where the tenant
has admin-consented it, and falls back to CLI/REST/SDK (
m365/az/pac/ PnP / Graph SDK / Power BI REST / Tabular Editor) otherwise./onboardprobes which is available. SeeCLAUDE.mdConnection policy. - Adaptive, role-driven
/onboard. A deeper interview that branches by department, maps the connections each person needs, drafts theirIT-REQUEST.md, and scaffolds starter skill stubs matched to their recurring work. No fixed persona ships — the kit is blank until you run it. - Microsoft-mapped intake + connections.
aios-intake.mdandconnections.mdpoint at Outlook, Teams, SharePoint/OneDrive, Business Central, Dataverse, Power BI, Power Platform, Azure. - Write guardrails for a real tenant. Read-only by default; any send/create/update/publish needs explicit confirmation; no autonomous sends. (The masterclass's 150k-inbox accident is why.)
- CLI/REST reference guides ship in
references/:business-central.md,ssrs.md,m365-cli.md,graph-rest.md,dataverse.md,powerbi-rest.md— plusonboarding/role-branches.md. - Worked examples in
examples/skills/— four draft-only report skills (Dispatch & Stores) you read and copy; they don't load as live commands.
Clone the repo, open it in Claude Code, run /onboard.
🆕 Setting this up on a new (work) computer? Follow
SETUP.mdstart to finish.
A free, MIT-licensed starter kit that turns Claude Code into your personal AI Operating System (AIOS). Audience: anyone building automations — solopreneurs, small business operators, managers, creators, AI consultants. Pairs with a companion masterclass video.
The kit personalizes itself to you via an /onboard interview, then gives you two recurring thinking skills (/audit, /level-up) to keep building leverage week over week.
AIS-OS stands for AI Automation Society OS — the way Nate designed this AIOS to be set up for members of his community, AI Automation Society. The kit is universal (it works for anyone), but the structure mirrors how AIS members run their own businesses on top of it.
"While you're not at your desk, your AIS-OS observes one real-world event and produces an output that's faster and more accurate than what you'd produce yourself."
Every design decision in this kit rolls up to that test. If a layer, skill, or template doesn't contribute to it, it doesn't ship.
Three felt success indicators tell you the AIOS is actually changing how you work. Not KPIs — there's no objective metric. These are lived experiences that show up in your week.
1. Team-reaches-out:
"A teammate messages you with a question. You realize your AIOS would answer it better, faster, and with exact sources — even if you were awake and free. So you ask your AIOS too. That's the moment you stop being a bottleneck for your own knowledge."
2. Context-switching reduction:
"You stop opening new tabs. You stop launching the desktop app. When something new lands, your first move is to ask the AIOS, not to open six things. The default surface for thought work shifts. Silent. Compounding."
3. Knowledge-leaves-your-head:
"You stop trying to remember business facts. You don't rehearse what you decided last quarter or what your customer said in that meeting. You trust the retrieval. The AIOS holds the truth, you hold the questions."
Personal foundation → company AI-readiness. Once these indicators show up for one person, the same data architecture powers everything else. Custom dashboards on the data you already collect. Automations on top of the connections you already wired. Team rollout where everyone has theirs. A company where every operator runs a personal AIOS is a company that's actually AI-ready.
The kit teaches personal AIOS first. Everything scales from there.
The kit teaches two complementary frameworks. Three Ms first, Four Cs second. Without the brain rewire, the architecture is just a folder structure.
| M | One-liner |
|---|---|
| Mindset | Default Shift, Function Breakdown, Curiosity Rule. To what extent can AI be leveraged here? |
| Method | Find Constraint → EAD (Eliminate, Automate, Delegate) → Map Process → Pick Autonomy Level → Tie to KPI. |
| Machine | Lego Principle, Validation Chain, Bike Method, Intern Rule, Kill Switch. Boring is beautiful. Workflows beat agents. |
Full breakdown in references/3ms-framework.md. The /level-up skill walks you through all three weekly.
The Three Ms of AI™ is a trademark of Nate Herk. © 2026 Nate Herk.
| # | Layer | One-liner | "This layer is in place" test |
|---|---|---|---|
| 1 | Context | Knows your business | Fresh Claude session answers "what does this business do and who works here?" without browsing |
| 2 | Connections | Reaches your stuff | "What's on my calendar tomorrow and what tasks are due?" → live data, no paste |
| 3 | Capabilities | Knows how to do the work | A short phrase triggers a multi-step workflow that produces an artifact |
| 4 | Cadence | Runs without being asked | Laptop closed. A brief lands in the inbox. A teammate messages it and gets a real answer |
Brand line: Context. Connections. Capabilities. Cadence.
The Four Cs of an AIOS™ is a trademark of Nate Herk. © 2026 Nate Herk.
Dependency graph: Context is non-skippable. Connections + Capabilities can build in parallel. Cadence is last — don't automate workflows that don't work manually.
The kit is intentionally lean. Skills here are ideation prompts and thinking tools, not heavy automations. You hack on top of the structure.
| Skill | Type | When to run |
|---|---|---|
/onboard |
Setup wizard (one-time) | Day 1, immediately after clone. Adaptive role-driven interview. Generates context + connections + IT request + starter skill stubs, and fills CLAUDE.md. |
/audit |
Recurring thinking skill | Day 7, then weekly. Four-Cs gap report. Read-only. Watch the score climb. |
/level-up |
Recurring thinking skill | Day 14, then weekly. Three Ms interview (Mindset → Method → Machine). One run = one shipped artifact. |
/project |
Workspace scaffolder | When a workstream needs its own scoped space. new / list / archive. Creates a sub-OS under projects/. |
/audit asks "is the AIOS built right?" (form). /level-up asks "what business leverage am I missing?" (function). They work in series — fix structure first, then capability planning becomes meaningful.
- Clone the repo to a working folder on your machine.
- Open it in Claude Code and run
/onboard. Answer the core questions, then your department branch. Voice samples must be pasted, not described. At the end you get your context files, a tailoredconnections.md, a filledIT-REQUEST.mdto hand to IT, and 2-3 starter skill stubs. - Hand
IT-REQUEST.mdto your IT team so the live-data path can be set up in parallel. You're productive off exports meanwhile. - Use it for a week. Bring real questions. Make real decisions. Log them in
decisions/log.md. - Day 7: run
/audit. Read the Four-Cs gap report. Pick one gap to close. - Day 14+: weekly
/level-upritual — matures one skill stub or ships one new automation per week.
aios-work/
├── README.md
├── CLAUDE.md ← Your operating manual (filled by /onboard)
├── SETUP.md ← Start-to-finish setup on a work PC
├── IT-REQUEST.md ← Entra-app / access request for IT (filled by /onboard)
├── EXPANSIONS.md ← What to add as you grow
├── LICENSE
├── .gitignore
├── aios-intake.md ← Source-of-truth for /onboard. Edit + re-run any time.
├── connections.md ← Registry of every system your AIOS can reach (MCP where consented, else CLI/REST)
├── context/ ← About you, your work (filled by /onboard)
├── projects/ ← Scoped workspaces (sub-OS per workstream), created by /project
│ └── README.md
├── references/
│ ├── 3ms-framework.md ← The operator brain
│ ├── onboarding/role-branches.md ← Department branch question sets for /onboard
│ └── *.md ← Microsoft CLI/REST guides (BC, SSRS, m365, Graph, Dataverse, Power BI)
├── examples/
│ ├── README.md
│ └── skills/ ← Reference skill patterns (NOT live commands)
├── decisions/
│ └── log.md ← Append-only record of what was decided and why
├── archives/ ← Old stuff. Don't delete. Move here.
└── .claude/
└── skills/
├── onboard/SKILL.md
├── audit/SKILL.md
├── level-up/SKILL.md
└── project/SKILL.md (+ your role skills, generated by /onboard)
See EXPANSIONS.md for what to add as you grow (projects/, templates/, scripts/, .claude/agents/, sub-OS folders, etc.).
MIT License. © 2026 Nate Herk.
The Three Ms of AI™ and The Four Cs of an AIOS™ are trademarks of Nate Herk. Both frameworks ship in this repo with attribution. Use freely; don't repackage as your own.
The companion masterclass video walks you through the kit step by step. Link will land here once it ships.