Back to Free Stuff

Free Guide — August 2026

How to Set Up Claude
in One Day

If you are a beginner with AI, your setup is the most important part of your system, and most people do it wrong. Five steps, about 30 minutes, done once. At the end of today Claude actually knows you and has something running for you.

Claude logo

Before You Start

Setup beats prompting, every time

Almost everyone starts in the wrong place. They go looking for the perfect prompt, the secret command, the cheat code someone posted. Then they wonder why the output still reads like generic AI slop.

Here is the thing nobody tells beginners: a great prompt on an unconfigured Claude loses to an average prompt on a configured one. Every single time. Because the configured one already knows who you are, what you are working on, what your writing sounds like, and it can reach into your real tools.

That is what today is. Not tricks. The foundation the tricks sit on. And you only do it once.

The Day

All 5 steps, about 30 minutes

Do them top to bottom. Each one builds on the one before it, so the order genuinely matters. You can stop after step 2 and still be ahead of most people, but the payoff is step 5.

1

Open Cowork instead of Chat

5 min

Everything else in the day happens in here. This is the one that makes the other four possible.

2

Tell Claude who you are

10 min

Paste the interview prompt. Claude asks you questions, then writes your global instructions. You paste those into settings.

3

Connect your tools

5 min

Gmail, Calendar, Drive, and anything else you actually use. This is what lets Claude do things instead of just talk about them.

4

Build your voice profile

5 min

Drop in 5 to 10 writing samples. Claude builds a profile it references every time it writes as you.

5

Have Claude build your first automation

5 min

It now knows you. Have it find the highest-leverage thing in your week and build it, running, before you close the laptop.

Step 1

Cowork, not Chat

If you are typing questions into a little chat box and copy-pasting the answers back into your work, you are using maybe 10% of what Claude can do. Chat answers you. Cowork works with you.

Cowork runs out of an actual folder on your computer. It reads your files, writes new ones, and remembers what it did last session. It connects to your tools and takes multi-step actions on its own, instead of you being the middleman between every step.

Download the desktop app at claude.ai/download if you have not already, then switch to Cowork and point it at a folder. Make a new one called my-ai-workspace if you do not have an obvious one. Do not run today from your phone.

Everything below happens in here.

Checkpoint

Ask Claude: "What files can you see in this folder?" It should list real files off your computer. If it cannot see anything, you are still in Chat.

Step 2 — Prompt 1

Tell Claude who you are

Paste this into a fresh chat. Claude interviews you one question at a time, like a friend getting to know you, then hands you a personal-instructions block. Answer honestly and do not rush it. This is the foundation everything else sits on.

When it gives you the block, copy it, go to Settings, find your global instructions (some versions call it Personalization or Profile), and paste it right in the box. Save.

From now on every new chat starts with Claude already knowing who you are. You just deleted the part where you re-explain yourself at the top of every conversation.

Copy-paste prompt

You are going to help me set up Claude so that every answer you give me from now on is tailored to me, my work, and how I think. We're going to do this as an interview, not a form. Do not skip ahead, and do not dump a giant list of questions at once.

The goal: by the end of this chat, you write me a personal-instructions block I can paste into Claude's settings, so every future conversation already knows who I am, what I do, what I'm trying to get done, and how I like you to respond.

Rules for the interview:
- Ask me ONE question at a time, then wait for my answer before the next one.
- Start broad, then go deeper based on what I say. Follow the thread.
- If an answer is vague, ask a follow-up until it's concrete. Push me for specifics: real numbers, real names, real examples.
- Keep it conversational, like a smart friend getting to know me. No survey energy.
- Roughly 12 to 18 questions total. Don't stop early just because you think you have enough.

Cover these areas, but in a natural order, not as a checklist you read out loud:
1. Who I am. Name, where I live, what I do for work, my role, how long I've done it.
2. My work. The business or job in detail. What I actually produce. Who my customers or team are. How I make money or get judged.
3. My goals. What I'm trying to hit in the next 90 days and the next year. What "winning" looks like for me.
4. My week. How I actually spend my time. What eats my hours. What I dread. What I'm best at.
5. My skills and gaps. What I'm great at, what I'm weak at, what I'd love to just hand off to you.
6. How I communicate. My tone, the words I use, the words I hate, examples of writing that sounds like me.
7. How I want you to respond. Long or short. Direct or gentle. Push back on me or agree. Ask before acting or just go.
8. Anything else a great assistant would kill to know about me.

When you have enough, say: "I have what I need. Here's your personal-instructions block." Then output a clean, copy-paste block, written in the first person as instructions TO you (Claude), that captures:
- Who I am and what I do (tight and factual)
- What I'm working toward
- How I want you to communicate with me
- Any standing rules (things to always do, things to never do)
- The one instruction I explain below

Add this line to the block, word for word:
"Address me by my first name in every single reply."

That's a reliability trick. The moment you stop using my name, it's my signal the chat has gotten too long and I should start a fresh one.

Keep the final block under 400 words. Plain language, no fluff. This is the thing I'm going to live with in every single chat, so make it count.

Start now with your first question. Just one.

Checkpoint

Open a brand new chat and say "hey." It should greet you by your first name and reference something about you. If it does not, the block did not save.

The name trick

Your block tells Claude to use your first name in every reply. That is not vanity. When Claude stops using your name, it is a free warning light that the chat has run too long and you should start a fresh one.

Step 3

Connect your tools

Out of the box, Claude can only talk. Connectors give it hands. Go to Settings, then Connectors (or Customize Connectors), and turn on the ones you actually use.

Start with these three and you have most of the value on day one.

Gmail

Inbox triage, drafted replies, follow-ups it writes in your voice once step 4 is done.

Calendar

Daily agendas, meeting prep, and a morning brief that actually knows your day.

Drive

Pulls context from your real docs and generates new files where you already work.

Not listed?

Google the tool name plus "MCP connector." It exists for roughly 95% of tools. If it truly does not, tell Claude which tool and it will give you the next-best path.

Checkpoint

Ask: "What is on my calendar tomorrow?" It should pull the real answer. If it offers to help you check, the connector is not live yet.

Step 4 — Prompt 2

Build your voice profile

Step 2 taught Claude who you are. This one teaches it how you sound. Those are different things, and skipping this is why most people's AI writing still reads like a press release.

Drop in 5 to 10 real writing samples: emails, texts, LinkedIn posts, Slack messages. Give it variety on purpose. The more different situations it sees you in, the better it understands how you shift between them.

This prompt does not just build the profile, it tests itself against you and rewrites until it actually sounds right.

Copy-paste prompt

I'm going to paste 5 to 10 samples of my real writing below. I want you to study them and build me a voice profile you reference every single time you write anything as me.

Give me variety on purpose. The more different situations you see me in, the better you understand how I actually shift between them. Pull from:
- 2 or 3 emails (one to a client or a boss, one to a friend or a teammate)
- 2 or 3 text messages or DMs
- 1 or 2 LinkedIn posts or captions
- 1 or 2 Slack messages
- Anything else I write regularly

Study all of it for:
- Sentence length and rhythm. Am I short and clipped, or long and winding? Does it change by context?
- The words I actually reach for, and the level I write at
- How I use punctuation. Em-dashes, parentheses, ellipses, lists, line breaks
- How I open. Do I lead with the point, warm up first, or start with a question?
- How I close. Sign-offs, CTAs, or do I just stop?
- Tone by context. How different am I in a client email versus a text to a friend?
- What I never do. Words I avoid, patterns that would immediately read as not-me

Then write me a voice profile with four parts:
1. THE RULES — 10 to 15 direct, followable rules ("Lead with the result." "Never use the word leverage." "Contractions always.")
2. CONTEXT SHIFTS — how the voice changes between formal and casual, and what stays constant in both
3. MY TELLS — the 3 to 5 specific things that make my writing recognizably mine
4. THE ANTI-LIST — words, phrases, and patterns to never use because they would blow the whole thing

Save it as voice-profile.md so any future chat can read it instead of relearning me from scratch.

Then prove it works. Write two short things in my voice: a reply to a client pushing back on a deadline, and a text to a friend cancelling plans. Show me both. If either one sounds off, I'll tell you what's wrong and you'll update voice-profile.md with what you learned. We repeat that until it sounds like me.

Here are my samples:

---
[Paste sample 1]
---
[Paste sample 2]
---
[Paste sample 3]
---
[Paste sample 4]
---
[Paste sample 5]
---

Checkpoint

Ask it to write a two-line text cancelling plans with a friend. If it sounds like a press release, feed it more samples and try again.

Step 5 — Prompt 3

Have Claude build your first automation

This is the one that surprises people, and it is the one most beginners overcomplicate. They think they need to find the perfect tutorial, or build some complicated workflow in another tool.

You do not. When you do not know how to do something with AI, just ask it to do it for you. AI understands itself better than you do. Explain in plain English what you want, ask how it would recommend building it, then tell it to go build it.

Claude now knows who you are, how you write, and what tools you have. That is exactly what it needs to find the highest-leverage thing in your week and build it, running, before you close the laptop.

Copy-paste prompt

You now know a lot about me: my personal instructions and my writing voice. I want you to use all of it to find the single highest-leverage thing in my life to automate, then actually build it.

Do this in four steps. Talk me through each one, don't just do it silently.

STEP 1 — REVIEW
Read everything you know about me: my personal instructions and my writing style. If anything looks thin, ask me a couple of quick questions to fill the gaps before you go further.

STEP 2 — FIND THE ONE
Look at all the recurring work in my life and find the single best task to automate right now. Score the candidates on three things:
- Time: how many hours a week or month it eats
- Dread: how much I hate doing it, or how often I drop it
- Doability: how cleanly you could actually do it with the tools I have
Show me your top 3 as a short ranked list with your reasoning, then recommend the ONE to build first. Let me confirm or override before you build anything.

STEP 3 — BUILD IT AS A SKILL
Once I confirm, build this automation as a reusable skill, so it runs the same way every time and I never have to re-explain it. Use the built-in Skill Creator to do it properly. Tell me plainly:
- What the skill does, step by step
- What it needs from me or my connected tools to run
- Anything I have to connect or turn on first, and exactly how to do it
Then build it and run it once as a test, so we both watch it work. Fix anything that's off.

STEP 4 — PUT IT ON AUTOPILOT
If this task should happen on a schedule (every morning, every Friday, the first of the month) or whenever something specific happens, set it up as a scheduled task so it runs without me. Tell me exactly when it'll run and what it'll do each time. If it genuinely shouldn't be automatic, say so, and tell me the trigger that should make me run it by hand.

One rule: build ONE thing, and build it well. Don't try to automate my whole life today. We get one real win running, I live with it for a week, then I come back and we build the next one.

Start with Step 1.

Checkpoint

Ask: "What skills do I have saved, and when does the scheduled one run next?" It should name your automation and the schedule.

That Is The Day

What to do tomorrow

You did the part almost nobody does. Claude knows who you are, sounds like you, can reach into your real tools, and has one thing running without you. That is a system, not a chatbot.

Three things to carry forward:

Live with the automation for a week before you build another one. One thing running beats five things half-built.

When you catch yourself explaining the same task to Claude twice, that is the signal to make it a skill.

Add to your second brain as you go. Every time Claude gets something wrong about you, that is a missing file, not a bad model.

You only had to do this once. Everything you build from here sits on top of it.

Work with Me

Want me on your problem for an hour?

Book a 1:1 call and we build the fix live — the workflow you want automated, the tool you can’t crack, whatever’s stuck. Direct, hands-on, no pitch waiting at the end.

$749 $349 credited back if we keep working together
See how it works

Work with Me

Want me on your problem for an hour?

Book a 1:1 call and we build the fix live — the workflow you want automated, the tool you can’t crack, or nothing pre-planned at all. No pitch waiting at the end.

See how it works

Full refund if we don’t find something worth building.