People ask me some version of the same question every couple of weeks: "I'm using ChatGPT a bit (one friend calls it Chad) and just got a new license at work, where do I actually start?"

I never had one good answer. I'd like to sit at a whiteboard with them in a room and help them step by step. So I made the thing I wish I could just hand them, and this issue is me handing it to you first.

Here's the core of it, and you can use it before you click anything.

The prompt

Most people use AI like a search engine. Ask, get an answer, close the tab, re-explain yourself tomorrow. Forget looking for a better model, the fix is giving the model a small amount of context that survives between sessions.

The prompt below does that for you. Paste it into Claude or ChatGPT, answer its questions for about five minutes, and it writes your setup files for you:

You're going to help me set up a reusable context profile for working
with AI tools. I want files I can reuse in every session so I don't
re-explain myself every time.

Interview me one question at a time. Do not ask everything at once. Cover:

1. My work — my role, what I'm responsible for, what I'm trying to get
   done over the next few months.
2. My tools — what I work in day to day (apps, file types, whatever's
   relevant to me).
3. How I like to work with AI — plan first or just do it, how blunt you
   should be, response length, anything that annoys me.
4. My recurring tasks — the things I'd hand to an AI every week if it
   already knew my context.
5. If I create content — who it's for, my tone, one or two things I'd
   say and one or two I never would. Skip this if it doesn't apply to me.

Ask a follow-up when an answer is vague. When you have enough, output
three files as separate Markdown code blocks:
- about-me.md — who I am and what I'm working on
- working-preferences.md — how I want you to collaborate with me
- brand-voice.md — only if I create content

Keep each file under ~200 words, in my words, no generic filler. Ask me
the first question now.

That's it. The interview makes it specific to you instead of generic, and the files are the part that compounds.

What you do with the output

You get three short text files. Where they go depends on the tool: Claude.ai takes them as "Project knowledge", ChatGPT takes them in "Custom Instructions", Claude Code reads a CLAUDE.md in your working folder. Same files, just stored in different paths (pro tip: the LLM will tell you where it needs to go). After that, every session starts with the AI already knowing who you are and how you like to work, instead of you typing it again.

The first time it answers as if it already knew you, you stop treating it like a search box. That's when you'll start seeing real benefits.

Read the full guide →

Getting Started with AI Tools: A Practitioner's Map

It's a 9-page do-it-with-me guide. Free to read, free to share. It walks the same setup for non-technical and technical paths, plus terminal and cloud agents when you're ready for them. The editable templates are free on GitHub too: github.com/cwilkins507/ai-adoption-playbook. If you'd rather have the whole AI Adoption Playbook (every template, an AI Readiness Worksheet, and the guide as a PDF) emailed to you, that's here.

If you want to go deeper

• "What is an AI agent harness?" — the mental model that connects all of this. Read this if the prompt above worked and you want to know why.
• "Context engineering for AI coding tools" — the four-layer version for people already writing code with AI.
• "My daily AI workflow" — what a fully configured day actually looks like.

Start with one thing you'd hand an AI every week if it already knew your context.

— Collin

If this was forwarded to you, you can subscribe here: buttondown.com/collinwilkins.