Hey future-proof friends 💜
So the top 1% of Claude users aren’t really wasting time writing long and detailed prompts anymore.
But personally, getting out of the 2025-style "paste a 500-word prompt every time you open a chat" habit has been a real challenge for me.
Turns out the clever way to do it is to set up your Claude Cowork properly once, so it gives you personalised, on-brand results without you having to specify what "good" looks like every single time.
So that's what we're diving into this week. A super simple Cowork setup that'll make your work 10x faster (because you're not repeating yourself) and your results 10x better.
But before we dive in:
100+ Claude Code hacks to ship code 10X faster
Top engineers at Anthropic and OpenAI say AI now writes 100% of their code.
If you're not using AI, you're spending 40 hours doing what they do in 4.
These 100+ Claude Code hacks fix that and help you ship 10x faster.
Sign up for The Code and get:
100+ Claude Code hacks used by top engineers — free
The Code newsletter — learn the latest AI tools, tips, and skills to code faster with AI in 5 minutes a day
One Serious Deep Dive 💡: Super simple way to set up your Claude Cowork the proper way in 30 minutes, so you can stop rewriting the same prompts.
Piping Hot AI Tea 🫖: Instagram just killed reach for reposted content, Google made AI video generation free for everyone, and AI just caught pancreatic cancer 3 years before doctors could.
💌 Your say genuinely shapes this newsletter: there’s a one‑click feedback poll at the very end. I genuinely check for feedback like a maniac after I send this out because I really want to know what you honestly think. So thank you!
One Serious Deep Dive 💡
Let’s set up your Claude Cowork the proper way, once and for all…
Part 1: Your Global Instructions (the "About Me" file)
Global Instructions are preferences Cowork applies to every session.
How to set it up

Click the settings icon at the bottom left of the Cowork window
Click Cowork in the settings menu
You'll see a field labeled Instructions for Claude
Paste in a version of the template below and customise it:
# Who I am
Name: [Your name]
What I do: [One-sentence description. Example: "I run a content business helping women 35+ use AI to build leverage and future-proof their careers."]
Brand: [Your brand name, if relevant]
Primary tools: [Claude, Notion, Google Drive, Canva, Wispr Flow, etc.]
# How I want you to work with me
- Always show me your plan before taking any action that touches files. I want to approve or redirect before you execute.
- Never delete anything without explicit confirmation.
- If something is unclear, ask me before assuming.
- Check in with me if a task is taking an unexpected direction.
- Be direct. Skip preamble. Don't over-explain unless I ask.
# What I'm working on
- [High-level context about your current projects, clients, or focus areas]
- [Example: "I'm actively producing short-form video content, a newsletter, and digital products."]
# What I'm sensitive about
- Never access folders or files I haven't explicitly given you access to.
- Don't touch anything that looks like credentials, API keys, or sensitive client information.
- When in doubt about scope, ask.
# How I like outputs
- [Example: "Write in my voice. I've set out my voice in the Claude.md for each folder, but in general: short punchy sentences, conversational, no corporate language, no hustle-culture language, no em dashes."]
- [Example: "For documents, deliver clean markdown unless I ask for something else."]Why this works
Every time you open a new Cowork task, Claude will load these instructions silently in the background. You'll immediately notice Claude behaves more like you want, without you having to repeat yourself in every prompt.
Part 2: Your folder system
Cowork works inside folders on your computer. One of the most common mistakes beginners make is pointing Cowork at their whole desktop and then wondering why it's slow, confused, or risky.
Instead, create a dedicated folder on your computer called Cowork.
Inside it, create a subfolder for each major project or area of your work. For example:
Cowork/
├── Content System/
├── Newsletter/
├── Client - [Name]/
├── Client - [Name]/
├── Personal Brand/
└── Admin/This will become Cowork's workspace. Every task you give it will happen inside one of these folders. Each folder stays separate, so the context from your client work never bleeds into your personal content, or across separate work projects.
Part 3: Your Claude.md files (folder instructions)
This is the really good bit… 🔥
Each folder can have its own instructions file called CLAUDE.md. When you point Cowork at that folder, it reads the instructions first and applies them to everything it does in that folder.
Think of Global Instructions as your "About Me" and CLAUDE.md files as your "About This Project."
How to create a Claude.md file
The easiest way is to have Claude create it for you. Open a regular Claude chat in your browser and say something like:
I'm setting up a Cowork folder for [project]. Can you write me a CLAUDE.md file with folder instructions?
The project is [describe it in detail, including:
- A short overview of the project
- The tone of voice you want Claude to use
- Brand principles and language to avoid
- The folder structure and what lives where
- Any rules you want Claude to follow when working in this folder
- What a great result looks like for typical tasks in this folder
- etc..]Claude will generate a markdown file. Download it or copy the output into a new text file named CLAUDE.md and save it in the relevant folder.
CLAUDE.md template for a content creator:
# [Project / Brand Name] Folder Instructions
## Overview
This folder is for [what you use it for]. Everything in here relates to [scope].
## Brand voice
- [Example: Short, punchy sentences]
- [Example: Conversational, slightly sarcastic]
- [Example: No corporate marketing language]
- [Example: No em dashes, no boss-babe language]
- [Example: Lowercase for subject lines and hooks]
## Folder structure
- /drafts — work in progress
- /final — completed work ready to publish
- /assets — images, logos, brand files
- /templates — reusable structures
## Rules for working in this folder
- Always save work in /drafts first, never overwrite files in /final without asking
- For any output that will be publicly shared, show it to me before saving to /final
- Name files using: YYYY-MM-DD-descriptive-name
## What "done" looks like
For [task type]: [what a complete deliverable looks like]
For [task type]: [what a complete deliverable looks like]💡 Pro tip: Once you have a CLAUDE.md template you like, save it somewhere central and adapt it for every new folder you set up. You'll create these faster and faster over time.
Permissions explained simply
The first time you point Cowork at a folder, it asks for permission. You have three choices:
Allow once - access for this session only. Good for one-off tasks or sensitive folders.
Always allow - access any time without asking again. Good for folders you use constantly.
Don't allow - cancel and rethink what you're doing.
Rule of thumb: Use Allow once for client work and anything sensitive. Use Always allow for your own brand folders and resources you'll return to again and again.
The safety checklist
Before you run your first task, run through this quick checklist:
I've created a dedicated Cowork folder, not pointed it at my whole desktop
I know what's in every folder I'm giving Cowork access to
Nothing in these folders contains passwords, financial account numbers, or private client data I can't afford to expose
I've backed up anything I can't afford to lose before letting Cowork edit it
My Global Instructions tell Cowork to ask before deleting
I'll read approval prompts carefully instead of clicking through them
⚠️ On sensitive data: Cowork runs on your machine and files stay local, but your prompts and task descriptions are sent to Anthropic. Don't put client data you're contractually required to protect into folders Cowork can see. If in doubt, leave it out.
Want to see how I use AI to 10x my content and audience growth?
I’ve created a done-for-you content ideation, creation, management and publishing system to help you leverage AI to 10x your content creation (without losing your authenticity in the process).
It’s called the AI Content Multiplier and the waitlist is now open - join now and I’ll send you a 30% off early bird discount as soon as it launches (which is soon).

Piping Hot AI Tea 🫖
1. Instagram Just Killed Reach for Reposted Content
Instagram has officially expanded its "unoriginal content" crackdown beyond Reels to include photos and carousels, slashing reach for any account that reposts content they didn't make. Reposted material is now largely excluded from Explore and recommendation feeds (which is the main discovery channel for growth). Aggregator accounts are reportedly haemorrhaging reach overnight, and original creators are finally being rewarded for, you know... actually creating. About time 👏
2. Google Just Made AI Video Generation Free for Everyone
Google rolled out Google Vids to the public last week, letting anyone with a Google account generate up to 10 AI videos a month at no cost. It's pitched as a Canva-killer for video, and for anyone making short-form content, slide decks, or promo material on a tiny budget... this is huge. Worth testing this week alongside ChatGPT Images 2.0 to see what fits your workflow.
3. AI Just Caught Pancreatic Cancer 3 Years Before Doctors Could
Mayo Clinic researchers revealed an AI model called REDMOD that can detect pancreatic cancer on routine CT scans up to three years before clinical diagnosis (even when the tumour isn't yet visible to the human eye). Pancreatic cancer is one of the deadliest cancers precisely because it's almost always caught too late. This is the kind of AI story we don't talk about enough. While everyone's panicking about job loss and deepfakes, AI models like this are about to save a lot of lives 💜
If you enjoyed today's newsletter AND got to the end of it, I’d love a quick click on the poll below to let me know what you think 💜.
See you next week,
Jess xx



