Claude
The One File that Controls How Smart Claude Is
Set up a CLAUDE.md file once, and stop re-briefing Claude on the same project every single conversation
Summary
A CLAUDE.md file is a markdown file that tells Claude how to operate on a project before it does anything — write it once, and every new conversation opens already knowing your context instead of starting from zero.
Before you start
Why this file is the most important file in your project
Before any conversation, Claude will read a file at the root of a project calledCLAUDE.md.
This is a markdown file, or .md that gives project-specific instructiosn and memory.
It's an automatic, onboarding manual that Claude automatically reads before you even start a conversation.
Create this right and Claude will know how your project works.
Get it wrong, or leave it empty, and you'll waste time re-explaining context Claude should already know.
Here's how to build a markdown file in 5 steps.
What is a markdown file, and why does Claude read it automatically?
A markdown file is just a text file. It's not special software, in no proprietary format.
What makes CLAUDE.md different from other files Claude reads is it will automatically read it at the start of a conversation.
Claude and other chatbots read many markdown files. But the file named CLAUDE.mdwill be the first file that it reads.
It should contain context to your project, universal instructions, and what "success" means to this project.
Caution: this only works if the file actually gets read. You must make sure this is saved within your Claude project folder or actual local computer folder (if you're using Claude Code).
Choose one project you're tired of re-explaining
Don't start with explaining your whole job.
Choose one project where you've caught yourself typing the same context.
Open a conversation and attach any files that are already relevant — a scope doc, a stakeholder list, or a status template.
You can also start with nothing and go straight to the interview in the next step.
These are the types of projects you could create a CLAUDE.md for:
A recurring client project where the stakeholder list and reporting cadence never change, but you type them out fresh every week
A cross-functional build where three teams have different definitions of "on track," and you keep re-stating whose definition wins
A campaign where you're referencing the same brand guidelines
Caution: picking a project that's still in early scoping, where the goals themselves are moving, will produce a file that's out of date in a week. Pick something with a stable shape, even if the details inside it are still active.
Let Claude interview you to build the file
Please don't write this file from scratch!
Paste the prompt below for Claude to interview you about the project.
"Interview me about [project name] so you can build a CLAUDE.md file for it. Ask me about scope, stakeholders, what's already decided, and how I want status updates written. Then draft the file from my answers."
Claude will prompt you with follow-up questions before it drafts a response.
You can select the options it populates for you, or write your own response. The file is only as useful as the specificity you give it.
Claude will then create a markdown file for you to preview.
Keep the file under 200 lines
This is the step most people get wrong, and it's the one that actually determines whether the file helps or hurts.
The instinct is to add more — more history, more edge cases, more detail — on the theory that more context means better output.
It doesn't work that way, for two reasons:
The longer the file, the longer Claude takes to read it before it can start on your actual request
An overly specific file over-constrains Claude's thinking. You'll get answers that follow your instructions to the letter and miss the better option you didn't think to write down
Treat it the way you'd treat onboarding a strong hire: give them the context they need to operate with judgment, not a script that removes the judgment entirely.
Caution: if your draft comes back over 200 lines, don't just delete sections at random. Cut the parts that describe history — decisions Claude doesn't need to remember — and keep the parts that describe how you want it to operate going forward.
Upload it to your project's context folder
Once you have the file, upload it into the project's context folder rather than pasting it into a single chat.
This is what makes it persistent. Every new conversation in that project reads it automatically, instead of only the one conversation you pasted it into.
After it's uploaded, Claude stops guessing at your context and starts operating inside it.
Skip this step, and every gain from the interview and the drafting disappears the moment you close the chat.
Caution: if you update how the project runs — a new stakeholder, a changed cadence — update the file too. An out-of-date CLAUDE.md is worse than no file, because it actively steers Claude toward the wrong assumptions instead of just lacking them.
Step | What to do | Where it happens |
|---|---|---|
1. What's a markdown file? | Understand it's a plain text file Claude reads automatically | No action. It's just context |
2. Choose one project and paste prompt | Pick the project you re-explain most often | New Claude conversation |
3. Answer the interview | Respond to Claude's interview questions | Same conversation |
4. Keep it under 200 lines | Cut history, keep operating instructions | Editing the draft |
5. Upload it | Save the file where Claude reads it automatically | Project's context folder |
"In what scenarios can I use a markdown file at work?"
By role:
Program and project managers: start with your highest-friction recurring program, not your whole portfolio.
Chiefs of staff / ops leads: use it for the stakeholder map and reporting format first — that's the part you retype most.
Anyone managing more than one active project: build one file per project, not one file trying to cover all of them.
"Isn't this the same as just writing a good prompt each time?"
No. A prompt is instructions for one conversation. A CLAUDE.md file is instructions for every conversation in that project, read before you type anything. The whole point is removing the re-typing, not writing a better version of it.
"My project is small — do I really need this?"
If you're only ever going to reference it once, no.
But most "small" projects turn out to be the ones you reopen every week for months. The file pays for itself the second time you'd have re-explained something, not the first.