Claude has limited memory between conversations.

Every session, it's like meeting a stranger. This is actually a feature, not a bug—if you know how to use it.

That's what Projects are for: persistent memory. A folder of markdown files that live in Claude Code. You write context once. You reference it forever.

Without Projects: You start a chat, paste the same 3 paragraphs of background context, explain your goals again, remind Claude what your brand voice is. Then you ask your question. Half your token budget is gone on setup.

With Projects: You create a "Brand Voice" file in your project folder. You create a "Product Roadmap" file.

You create a "Customer Interviews" file. When you need Claude's help, you just reference them:

"Using my Brand Voice file and the latest customer interviews, write a tweet about our new feature." Projects turn Claude from a stateless tool into a collaborator that understands your world.

The best part: you can share projects with others. Your team can reference the same context. Everyone's on the same page.

Stop copying-and-pasting context into every conversation. Put it in a Project. Reference it. Move faster.

**What's one piece of context you find yourself pasting into conversations repeatedly?

That's your first Project file.*