CLAUDE.md is a markdown file in your project root that Claude Code reads automatically at session start, providing persistent memory about your stack, conventions, and preferences so you don't have to re-explain context every time.

Should I commit CLAUDE.md to version control?

Yes! The project-level CLAUDE.md (in your repo root) should be committed and shared with your team. This ensures everyone gets the same context when working with Claude Code.

How is this different from using /init?

The /init command in Claude Code analyzes your existing codebase to generate CLAUDE.md. This generator lets you create one from scratch by describing your project, which is useful for new projects or when you want more control over the content.

What should I include in CLAUDE.md?

Include your tech stack, development commands, code style conventions, architecture decisions, and project structure. Focus on universally applicable information that Claude should remember in every session.

Generator

CLAUDE.md Generator

Persistent memory · Claude Code onboarding

Free CLAUDE.md generator — what you get

This CLAUDE.md generator creates a persistent memory file that gives Claude Code context about your project. Place it in your repo root to eliminate repetitive explanations about your stack, conventions, and commands. Read more in our complete CLAUDE.md guide.

Generate a CLAUDE.md file that gives Claude Code persistent memory about your project. This onboarding script helps Claude remember your stack, conventions, and commands across sessions.

Project name *

Tech stack *

0 characters (min 10)

Development commands (optional)

Code style & conventions (optional)

Architecture decisions (optional)

Project structure (optional)

Additional context (optional)

FAQ

What is CLAUDE.md in one sentence?: CLAUDE.md is a markdown file in your project root that Claude Code reads automatically at session start, providing persistent memory about your stack, conventions, and preferences so you don't have to re-explain context every time.
Should I commit CLAUDE.md to version control?: Yes! The project-level CLAUDE.md (in your repo root) should be committed and shared with your team. The user-level CLAUDE.md (in ~/.claude/) is personal and should NOT be committed.
How is this different from using /init?: The /init command in Claude Code analyzes your existing codebase to generate CLAUDE.md. This generator lets you create one from scratch by describing your project, which is useful for new projects or when you want more control over the content.
What should I include in CLAUDE.md?: Include your tech stack, development commands, code style conventions, architecture decisions, and project structure. Focus on universally applicable information that Claude should remember in every session. Keep it under 150 lines for best results.
Can I have multiple CLAUDE.md files?: Yes! You can have a project-level CLAUDE.md in your repo root (shared with team) and a user-level CLAUDE.md in ~/.claude/ (personal preferences). Claude Code reads both, with project-level taking precedence.
How do I update CLAUDE.md over time?: CLAUDE.md is living documentation. When you adopt new libraries, change architecture decisions, or add conventions, update CLAUDE.md immediately. You can also tell Claude "save this to memory" during sessions to have it update the file for you.

Related resources

→What is CLAUDE.md? Complete guide to persistent memory →MEMORY.md: The long-term brain for AI agents →DESIGN.md: Open spec for AI design systems →What are agent skills? A complete guide