I keep seeing the same thing happen. Someone opens Claude, gets hit with words like Artifact, Connector, and Skill, and quietly closes the tab. It feels like everyone else got a manual you never received.
That’s why I got excited when I found this post. The author, an AI professional who’s clearly tired of the overtechnical noise, put together a plain-English one-pager on how Claude actually works. No jargon soup. Just the words you need and the exact steps to set it up. I was genuinely relieved reading it, because this contributor explains it the way a friend would, not the way a whitepaper does.
Let me walk you through what the creator shared.
First, the words nobody bothers to define
The original poster starts with vocabulary, because people throw these terms around like you’re supposed to already know them. Here’s the clean breakdown:
- Project: a folder with your uploaded files, rules, and sorted chats for one job.
- Style: a copy of how you write.
- Artifact: a doc that opens on the right side of your screen, which you can edit right there.
- Skill: a long prompt you wrote down once so you never have to explain it again.
- Connector: your tools inside Claude (Gmail, Drive, Slack) so it reads them and gets your context.
- Claude Cowork: it works on files stored on your desktop while you go do something else.
- Token: literally a word. That’s all it means.
Wasn’t that scary, right? That’s the whole point the expert makes. The words sound intimidating, but the ideas behind them are simple.
Now the setup, step by step
This is where the post gets practical. The creator lays out a clear process, and each step has a reason behind it. Here’s how it goes.
- Open a Project and name it after your repetitive task. Something like “Client emails.” Then upload every file tied to that task. Once you do this, Claude reads everything in the folder before it answers you, so it stops guessing and starts working with your real context.
- Write your about-me file and turn it into a Skill. The author says it should cover your job, who you sell to, what you sell, how you talk, and what you never say. That way Claude sounds like you instead of a generic robot.
- Put your rules in the Project instructions. This is the guardrail that keeps every answer on-brand without you repeating yourself.
For step two, the original poster shares the exact prompt to convert your file. Upload your .md file in a chat and paste this:
“Turn this uploaded .md file into a Claude skill with the name of ‘about me.'”
For step three, the expert gives a fill-in-the-blank template you can copy and adjust. Just swap in your own names and words:
“You help me write [what]. My readers are [who]. Use short sentences. Never use the words [x] or [y]. If you need more information, ask me one question at a time. When I paste a draft, tell me what to cut before you rewrite anything.”
I love that last line. Telling Claude to point out what to cut before it rewrites anything is a small trick that saves you from bloated drafts. That’s the kind of detail you only get from someone who actually uses the tool daily.
The features this pro deliberately skips
Here’s the part that surprised me most. The person who posted it doesn’t use everything Claude offers, and the reasoning is sharp.
- Memory stays off. The author argues it poisons future chats and boxes Claude in. Every conversation starts clean instead of dragging old assumptions along.
- Only three Connectors, max. This industry pro keeps it tight and warns you to never paste logins, keys, or client contracts into the tool.
Why it matters: more connections and more memory sound powerful, but the creator’s point is that restraint keeps Claude flexible and your data safe. Less clutter, better output.
My quick take
What makes this so useful is the order of operations. The mind behind it doesn’t just list features, they show you the exact sequence: define the words, build the Project, bottle your voice into a Skill, then set the rules. Follow that path and Claude stops feeling like a mystery and starts feeling like a teammate who already knows your job.
If you’ve been avoiding these features because the vocabulary scared you off, this is your permission to jump in. Start with one Project named after your most repetitive task. That single move does most of the heavy lifting.
The original post has even more nuance in how the expert phrases each step, so go read the full breakdown on LinkedIn and set your own Claude up today.