Claude in
5 Minutes.
Why this exists.
01 · WelcomeMost "intro to Claude" pieces are long. They walk you through Projects, Artifacts, Skills, and the philosophy of working with AI. By the time you've finished, your meeting started ten minutes ago and you still haven't done anything in Claude.
This one is different. You're going to do one useful thing in Claude today, in the next five minutes, with no setup beyond opening the app.
That's it.
What Claude actually is, in one sentence.
02 · Mental modelClaude is an AI you can talk to like a smart coworker, except it never gets annoyed when you ask the same question twice.
You don't have to write it special instructions. You don't have to "prompt engineer." For the next five minutes, just type the way you'd talk.
Open Claude.
03 · The doorwayIf you haven't already, go to claude.ai and sign in with your INFLXD email. If you have the desktop app installed, open that instead.
You're looking at a chat box. That's the whole interface. It's supposed to feel boring.
The one thing to do first.
04 · Your first useful promptFind the longest email thread or meeting doc you've been avoiding. The one you keep scrolling past because reading it feels like a chore.
Paste the whole thing into Claude and type:
"What's the action I need to take from this?"
That's the prompt. No formatting tricks, no role-setting. Just that question.
Claude will read the whole thing in seconds and tell you what you need to do, who you need to reply to, and what's safe to ignore.
Why this works better than "summarize this."
05 · The reframe"Summarize this" gives you a smaller version of the thing you already didn't want to read.
"What's the action I need to take" gives you a decision. It's the actual reason you opened the thread in the first place.
That tiny reframe is the whole trick. The more specific your ask, the more useful the answer.
Common questions before you start.
06 · FAQThese are the things people wonder about right before their first session. Short answers, no hedging.
Is Claude reading everything I paste, forever?
No. By default, the things you type and paste into Claude are not used to train future models, and your chats live in your account, not in a shared brain. You can delete a chat and it's gone. For real client-sensitive material, check INFLXD's policy with James or Fiona before pasting.
Will Claude make stuff up?
Sometimes. AI models can sound confident while being wrong. The fix isn't to stop using them, it's to treat the output the same way you'd treat a smart intern's first draft. Useful, fast, needs a sanity check. Never paste a Claude answer into something high-stakes without reading it.
Is this safe for client work?
For most internal drafting and thinking, yes. For anything that includes confidential client data, follow INFLXD's data-handling guidance. If you're not sure, that's the signal to ask before pasting, not after.
How is this different from ChatGPT?
Day to day, less than you'd think. Claude tends to be better at long documents, writing in your voice, and handling nuance. ChatGPT tends to be more aggressive about giving you a full answer fast. Both are useful. INFLXD's default tool is Claude, so this is the one to learn first.
Does Claude remember me between chats?
By default, no. Each new chat starts fresh. There are ways to give Claude persistent memory (see the Prompts That Pull Their Weight preread when you're ready), but for your first week, assume every chat is a clean slate.
Can I trust the answer?
Trust it the way you'd trust a smart, fast colleague who's never been to INFLXD. They can help you think. They can help you draft. They will get specifics wrong sometimes. The work is yours to ship; Claude is just helping.
What if Claude says something weird or wrong?
Tell it. Type "that's not right, here's what I actually meant" and try again. Most of the time it course-corrects immediately.
What to try when you have a spare minute later.
07 · Next movesOnce the first thing works, try these. Each one takes under two minutes.
- Decode a confusing message. Paste a Slack thread or email you didn't fully follow and ask "what is this person actually asking me?"
- Rewrite something you didn't want to write. Paste your rough draft of a status update or message and ask "make this clearer, keep my voice." Then edit.
- Get unstuck on a decision. Type "I'm trying to decide between X and Y. Ask me three questions that would help me figure it out." Let Claude interview you.
- Catch yourself up on a thread you missed. Paste a long Slack thread or meeting transcript and ask "I missed this. Catch me up in three bullets."
You don't need to do these today. Just know they exist.
What to skip (for now).
08 · Defer thisYou will see references to Projects, Artifacts, Skills, Cowork, MCPs, Claude Code. Ignore them all for now. They're useful, and you'll get to them. None of them are the first thing.
The first thing is the chat box. Use it for a week before you go looking for anything else.
If you only remember one thing.
09 · The whole pieceOpen Claude. Paste the thing you've been avoiding. Ask "what's the action I need to take."
That's the entire piece.
Where to go next.
10 · The bigger map- When you want a bigger picture of what Claude can do: read the Claude Preread on the resources page.
- When you want to actually delegate work, not just ask questions: that's Cowork, and there's a CoworkOS onboarding deck on the resources page.
- When you want your prompts to do more work for fewer tokens: read Prompts That Pull Their Weight on the resources page.
Sources.
11 · Worth keeping on the shelfThe framing in this preread draws on a few public pieces worth keeping on your shelf.
- Get started with Claude · Anthropic Help Center The official "open Claude, do this" doc.
- Claude 101 · Anthropic Academy Free, modular, 2-3 hours total if you want the full ramp later.
- Working with AI: Two paths to prompting · Ethan Mollick The canonical "just talk to it" piece.
- Everyone should be using Claude Code more · Lenny Rachitsky 50 ways non-technical people use AI tools day to day.