<!--
ROBOTS ON PAYROLL / THE VAULT
What this is: a beginner-friendly CLAUDE.md, the house-rules file Claude Code reads at the start of every session.
How to use it: download this file and give it to your AI (Claude, ChatGPT,
Cursor, Lovable chat) as a reference. Say: "Use this as my playbook for [task]."
Latest version + full guide: https://robotsonpayroll.com/resources/workspace-setup-kit
-->

<!-- HOW TO USE: rename this file to CLAUDE.md and put a copy in every
     project folder. Claude Code reads it automatically when it starts,
     so these rules apply without you repeating them every session.
     Delete the comment blocks (the parts that look like this one) when
     you understand them; only the plain lines are the actual rules.
     When you outgrow this file, the PRO version lives at
     robotsonpayroll.com/resources/claude-md-starter -->

# CLAUDE.md

I'm new to coding. I build by describing what I want in plain English.

<!-- Why: this one sentence changes how Claude talks to you all session.
     It explains more, assumes less, and skips the jargon. -->

## How to work with me

- After every step, tell me in one or two plain sentences what you changed and why.
- Make one change at a time. Finish it, show me, then move to the next thing.
- Explain any technical word the first time you use it, in one short phrase.

<!-- Why: the #1 beginner failure mode is the AI making 12 changes at once.
     When something breaks, you can't tell which change did it. One change
     at a time means every problem has exactly one suspect. -->

## Safety rules

- Ask me before deleting any file, always. Even ones you created.
- Never touch the .env file without telling me first. (That file holds secrets like passwords and API keys.)
- If a command could cost money or send something to the internet, warn me before running it.

<!-- Why: these three cover the only mistakes you can't easily undo:
     lost files, leaked secrets, and surprise bills. Everything else
     in coding is reversible. -->

## Keep a diary

- Keep a file called PROGRESS.md in this project. After each work session, add 2-4 bullet points: what we built, what broke, what's next.

<!-- Why: Claude forgets everything between sessions. PROGRESS.md is its
     memory AND yours. Next session, start with "read PROGRESS.md" and
     it picks up exactly where you left off. -->

## When something breaks

- Explain the error to me in plain English before fixing it. I want to learn what happened, not just have it disappear.

<!-- Why: each error you understand is one you'll fix yourself next time.
     Three months of this and you won't need the training wheels. -->