How I built an AI playbook for clear communication

Have you ever tried to delegate something, only to think: never mind, it’s faster if I just do it myself?

That happens when the instructions live only in your head. A playbook writes them down once, so handing off a task, to a person or to AI, stops being slower than doing it yourself.

What is an AI playbook

Think of it as an employee handbook for AI. You write the task down once, and AI follows it like a new hire would.

Every playbook has four parts:

  • Trigger: what starts it
  • Inputs: what AI needs from you each time
  • Steps: how you do the work, broken into tasks
  • Outputs: what AI hands back to you

You won’t get a playbook right the first time. Write the steps, run it, and note where it misses: too formal, too long, the wrong detail. Then adjust until it’s accurate enough, for now, for the job you built it for.

You don’t need a dedicated platform. Write the playbook as plain text, then run it anywhere: a chatbot, an agent, a multi-agent system, a Custom GPT, or a Project.


I built Master Your Message (MYM), an AI playbook for a task I do every day: communicating with stakeholders. It started as my project for Rachel Woods’ Think Like an AI Operator Challenge.


Why Master Your Message

If you lead cross-functional work, you know each stakeholder needs context explained differently.

I kept doing two things that slowed me down:

  • Burying key points in long paragraphs
  • Leaving my ask unclear

I kept rewriting the same update for different stakeholders. Engineers wanted technical details upfront. Marketing needed business impact. Leadership just wanted the bottom line.

I knew what I needed to say. I struggled to structure it for each audience.

MYM stops that rewriting. I answer three questions, and it gives me the right structure for that audience on the first try.


How Master Your Message works

MYM follows the same trigger, inputs, steps, and outputs shape as any playbook. It asks who you’re talking to, what you need from them, and how you’re sending it. It pushes back if the message is sensitive or outside what a playbook should write.

Then it matches your answers to a structure: Bottom Line Up Front (BLUF) for executives, or Point-Reason-Example-Point (PREP) for technical recommendations. It drafts the message and checks that the ask appears in the first line or two before handing it back.

You get a message ready to send, plus the framework it chose and why, so you don’t have to make the same call twice.


Master Your Message capabilities

MYM adapts to how much you give it. With a full brief, it drafts straight away and names the framework. With a vague one, like an API migration update with no audience, it asks questions before writing a word. When it can fill a gap itself, it tells you what it assumed.


MYM came from a task I was already repeating every day. To find your own playbook, look at your standard operating procedures, calendar, to-do list, project plans, and wishlist. Any task you keep redoing by hand is a candidate.