The setup that makes Claude yours

Stop chatting, start coworking

June 15, 2026

In partnership with

For months I assumed the people pulling magic out of Claude were just better at prompting than me. Turns out the gap was never about clever wording. It was about setup.

This one came from a post by an AI professional who shared a one-time configuration that turns Claude from a stranger into a teammate. The idea that hit me: the slowest way to use Claude is re-explaining yourself every single session. No theory, no fluff, just a folder, three files, and a handful of habits.

The GTM bets that shouldn't have worked, and did

One grew revenue 50x after half his team quit over the strategy. One brought in 50K signups in a single day with no paid budget. One generated 100M+ views from a stunt that took 50 hours to conceive. One asked every prospect to demo the product themselves instead of demoing it for them.

None of them followed the safe playbook. They treated GTM like an experiment, moved before they had proof, and made bets most founders would never get approved.

HubSpot for Startups documented all 6 stories in the free Bold Bets Playbook. The risks they took, why it was risky, and what it returned.

*Ad

Stop Chatting, Start Coworking

The creator's first point is the one most people skip. Regular Chat forgets you: it starts from zero on every conversation, so you keep re-explaining who you are and what you want. Cowork reads your files before it answers.

That single switch changes the math. You stop training the model from scratch, and the quality jump lands fast. Here's why I think it matters: context beats cleverness every time.

The Folder Is The Brain

This is the part I found genuinely smart. The author says a folder beats a clever prompt, and the structure won me over. Make one folder called Claude Cowork, add three subfolders: about-me, outputs, and templates.

Point Cowork at it and that folder becomes the model's memory. Then fix the defaults that quietly sabotage you: pick Opus 4.8 for hard tasks, turn on Thinking, set Effort to High.

The reasoning is simple. Light defaults are tuned for speed, not depth: flip on high-effort thinking and Claude reasons through the problem instead of rushing to a surface answer. Twenty minutes, once.

Payroll errors cost more than you think

While many businesses are solving problems at lightspeed, their payroll systems seem to stay stuck in the past. Deel's free Payroll Toolkit shows you what's actually changing in payroll this year, which problems hit first, and how to fix them before they cost you. Because new compliance rules, AI automation, and multi-country remote teams are all colliding at once.

Check out the free Deel Payroll Toolkit today and get a step-by-step roadmap to modernize operations, reduce manual work, and build a payroll strategy that scales with confidence.

*Ad

Three Files And Better Habits

Here's the move I'm stealing from this contributor: don't write your files from scratch. Open a fresh session, ask Claude to interview you, then build three: about-me, anti-ai-writing-style, and my-company.

Trim each one under 2,000 tokens. If a line doesn't change how Claude responds, cut it. Three sharp files beat fifty messy ones.

Then the habits, where the original poster gets real: the tools change nothing if your behavior stays the same. End prompts with 'ask me questions,' correct wrong answers instead of accepting them, and turn your best output into a reusable template. If you want the exact file structure, the full breakdown is worth the few minutes.

This is a smart, low-effort approach for anyone who wants Claude to work like a coworker instead of a vending machine. If you want to see the full breakdown, check out the original article.

Credits to the original Lin

One AI employee. Engineering, finance, growth, ops.

Last week Viktor opened 14 pull requests, closed two month-end books, drafted a board update, deployed three landing pages, and triaged 600 support tickets. From inside Slack and Microsoft Teams. 20,000+ teams now run this way.

*Ad