The Chief of Staff You Can't Afford to Hire

Every morning you do a job that has a title in bigger companies. Before the day starts, someone there reads across the calendar, the inbox, the messages, the deals, and the meeting notes, decides what actually matters, and hands the boss a one-page brief: here's what today is about, here's the first move. That someone is a chief of staff. In a company your size, that someone is you — half-awake, rebuilding your own to-do list from five open tabs and last night's worry.

A real chief of staff runs about $150,000 a year. You're not hiring one to run a six-person shop. So the deciding-what-matters job stays on you, and it gets done at 6am, badly, before coffee. That's not a discipline problem. It's a math problem — the one hour of the day that would most repay a clear head is the hour you have the least of one.

The most valuable hour of your day is the one you spend deciding what the day is about — and it's the hour you're least equipped to spend well.

The difference between an assistant and a chief of staff

This matters because it changes what you're building. An assistant screens and schedules — it hands you a sorted list of unread email. A chief of staff synthesizes: it reads everything, decides the two things that move your week, and drafts the first one before you ask. The first is a filter. The second takes judgment. And judgment is exactly the thing AI got good enough at this year to be useful for — if you point it at your actual work and keep a human on the trigger.

There's a version of this build going around right now — people wiring up a scheduled agent that posts a ranked morning brief to their Slack before they're out of bed. It's a genuinely good idea. The catch is that every write-up assumes you'll happily sit down and fill in a two-page system prompt by hand, naming your tools and your VIPs and your tone in exactly the right places. That's the part that stops a busy operator cold. So I did what this whole site is about: I turned the hard part into a paste.

You describe the business; it writes the staff

The free tool asks you a few plain-English questions. What's the business. What do you personally run day to day. Which tools do you actually live in. Who can never get buried. How do you want to be talked to. Then it hands back a finished system prompt — your chief of staff, written for your shop — that you paste into a Claude Routine and schedule for weekday mornings. It reads only the tools you named. It ranks instead of dumps. And it ends every brief with one concrete first move, so you don't spend your clearest thinking deciding where to start.

Read that last line again, because it's the whole point. The brief doesn't just tell you what to do. It offers to do it — a specific draft, with a time estimate — and then it waits. Nothing sends on its own. Every connector is read-only unless you say otherwise. The agent reads and decides what matters; you decide what happens. That's the "Automate" step of the READY framework done the safe way: real leverage, hands still on the wheel.

What it touches, plainly

The tool that writes your prompt connects to nothing. You type answers, they're sent once to an AI model to draft the prompt, the result shows in your browser, and nothing is stored. The connectors come later, and they're yours — set inside your own Claude, read-only by default, pointed only at the tools you chose. And by week three the brief starts earning its keep in a way a fresh assistant never could: it keeps a memory file, so it notices the follow-up you keep rolling and the VIP you keep meaning to call, and it mentions the pattern once. Not the brief. The behavior change.

You already do the chief-of-staff job every morning. This just gives you one that clocks in before you do.