It's been a while since my last issue and the reason is that my cat Blue has been going through some major health problems, and the last several weeks have just been a lot. What I didn't expect was that it would lead me to a workflow shift in how I use Claude. And before you roll your eyes, I promise this isn't one of those LinkedIn posts where someone is trying to force a connection between their personal life and work! Let me explain…

I had been using Claude pretty heavily during that time to keep track of everything, do research, and help me process vet conversations and results as they came in. What I noticed was that Claude wasn't automatically pulling all of the context from other threads within the same project like I thought it would.
I noticed a huge improvement when I gave it instructions within the project folder. You may be thinking duhhh Brittany… but this didn’t always work well in ChatGPT.

Once I figured out I could use the instructions to tell it to review previous messages before responding, I noticed a huge improvement. Every new thread it would orient itself to where we were, and I didn't have to re-explain the situation. At a time when I really needed that, it made a genuine difference!

And that's when I started thinking about how to bring this into my client work.

I had already been using Projects in a pretty surface-level way before this, basically just as folder organization so my threads for work, side projects, and personal stuff weren't all tangled together. But the real shift was realizing that a Claude Project can actually function like a living, breathing project brief if you set it up right.

Here's how I'm setting up client projects now.

I built a template framework with Claude's help, explaining what I use client projects for, what kinds of questions I'll be bringing to it, what to always prioritize, and what to avoid. Think of it as writing a creative brief, but the brief is for your AI collaborator. I've been refining this into a reusable template that I can adapt for each new client. Want to use it? It’s available in the playground section below!

At Wizardly, we have client profiles that house key information about each account: brand voice, strategy notes, scope of work, and things like that. I upload those into the project so that when I'm in the middle of a deliverable and I need a quick answer I can just ask and it can pull from the documentation rather than me stopping to dig through folders. It's like having a very well-briefed collaborator on call.

When transcripts from client calls get added to the project, the whole context gets richer over time. I’ve been thankful that I included every client transcripts, even workshops, into the folders because Claude has pushed back on ideas I had using direct information from the client in these transcripts!

This one I cannot stress enough. The setup saves time, but the output still needs you. I always go back through anything Claude helps build for a project and make sure it actually sounds right and reflects the nuance that only comes from being in the work. The AI is the bridge but it still lacks that human touch!

If you want to try this yourself, the best place to start is with one active project, whether that's a client retainer, a side project, or something personal. The setup doesn't have to be complicated, and honestly the instructions don't need to be perfect on day one. They just need to give Claude enough context to orient itself before it responds.

To make it easier to get started, I'm sharing the framework template I built for my own client projects. It's stripped of any specific client info so you can drop it in and make it your own. The most important line in the whole thing is in the "How to work with me" section, and it's the thing that made the biggest difference for me: telling Claude to always review previous messages before responding. That one instruction is what turns a folder of threads into something that actually has memory.

What this project is: [1–2 sentences on who this is for and what you're working on. Example: "Acme Co. is a B2B software company. We're building their marketing site and brand collateral."]

I'm a [designer / strategist / freelancer — fill in your role] working on this project. I handle [copywriting / design / brand strategy — whatever applies]. I am not a domain expert in the client's industry, so always defer to the documents in this project for positioning and terminology.

You're a strategic creative partner helping me produce polished, client-ready work. You understand B2B marketing, brand voice, and agency or freelance workflows.

[List any confirmed decisions here. Examples: "No em dashes in any copy." "Always use sentence case for headlines." Add to this as the project evolves.]

Power users figured this out early: speaking a prompt gives you 10x more context in half the time. You include the edge cases, the examples, the tone you want — because talking is fast enough that you don't skip them.

Wispr Flow captures everything you say and turns it into clean, structured text for any AI tool. Speak messy. Get polished input. Paste into ChatGPT, Claude, Cursor, or wherever you work.

89% of messages sent with zero edits. 4x faster than typing. Works system-wide on Mac, Windows, and iPhone.

Start flowing free