Cursor Rules and CLAUDE.md Generator
Choose your stack, set a few preferences, and copy a short project instruction file for Claude Code, Cursor, or any agent that reads AGENTS.md. You edit it before you use it.
Pick the file
Select the file your coding agent reads.
Save it as CLAUDE.md in your project root. Claude Code reads it at the start of a session. Jump to the result
Choose your stack
We use these choices to suggest conventions and commands. You can edit the commands below.
Language · Choose one
Framework · Choose one
Package manager · Choose one
Styling · Choose one
Testing · Choose one
Add technologies not listed above, one per line. They appear in the Stack section.
Describe your project
Give the agent a little context. Both fields are optional.
Set your preferences
Choose the rules you want the agent to follow every time. 4 selected.
Add your own preferences, one per line. They appear as rules in the file.
Check commands and add rules
These are suggestions for your stack, not a scan of your project. Edit them to match your scripts or leave a field blank to omit it.
Your CLAUDE.md
# Project ## Stack - TypeScript - Next.js - Tailwind CSS - Vitest for tests - npm as package manager ## Commands - Install: `npm install` - Run: `npm run dev` - Build: `npm run build` - Test: `npm test` - Lint: `npm run lint` ## Rules - Use the App Router. Prefer Server Components and add `use client` only where state or browser APIs are needed. - Check the installed Next.js version's documentation before using an API that may have changed. - Style with Tailwind utility classes. Do not add separate CSS files for things Tailwind can express. - Make small, focused changes. Do not refactor unrelated code. - Ask before adding a new dependency. - Add or update tests for every behavior change, and run the test command before finishing. - Never commit secrets, and do not print values from .env files.
Save it as CLAUDE.md in your project root. Claude Code reads it at the start of a session. Review the commands and rules before adding this file to your project.
What goes in a project rules file
A rules file tells a coding agent how your project works before it touches the code: the stack, the commands that run and check it, and a few rules it should follow every time. Claude Code reads CLAUDE.md, Cursor reads rules in .cursor/rules, and several agents read AGENTS.md. The three are the same idea in different file names.
Keep it short. The file is added to context on every task, so write down what the agent cannot guess from the code. Pair it with an llms.txt file for your public site, and use the AI crawler robots.txt generator to control which crawlers may read it.
FAQ
- What is the difference between CLAUDE.md, a Cursor rule, and AGENTS.md?
- They are all plain Markdown instruction files for coding agents. CLAUDE.md is read by Claude Code, Cursor reads rules from .cursor/rules/*.mdc (which add a small frontmatter block), and AGENTS.md is a shared file name several agents read. The tool writes the same rules in whichever format you pick.
- How long should the file be?
- Short. The file is added to the agent's context on every task, so keep it to commands, conventions, and rules the agent cannot guess from the code. This tool keeps the output compact on purpose.
- Are the default commands right for my project?
- They are common defaults for the stack you pick, not a scan of your repo. Edit any command to match your scripts before you copy the file.
- Is anything uploaded?
- No. The file is built in your browser from what you type and pick.