How to document an app you built with Cursor
Last updated: July 26, 2026
To document an app built with Cursor, point Explain My Build at your repo or local folder. Because Cursor is an AI IDE, you already have real code on your machine. It reads that code and writes an end-user guide, a founder runbook and a code-verified fix-list, each claim cited to a real file.
What is Cursor, and what do you already have?
Cursor is an AI-native code editor — a fork of VS Code rebuilt around AI — so unlike browser app builders, you already have a normal project on your own machine. Cursor-specific facts worth knowing:
- You own a real repo or folder. There's no export step; your code is right there in your file system, ready to read.
- Cursor's Agent and Composer wrote across many files. Composer makes multi-file edits and Agent mode can autonomously create, edit and delete files across your whole project. That's powerful — and it's exactly why you may not follow everything it built.
- It indexes your whole codebase. Cursor works from a full index of your project, which is why a single prompt can touch code in places you never opened.
So the documentation problem with Cursor isn't "where's my code" — it's "the AI built features across files I've never read, and I can't explain them to my users." That's what reading the code solves.
How do I document features Cursor built that I don't follow?
Point Explain My Build at the folder and let it read what Cursor's Agent produced. Because it reads the actual source, it describes every route, form and setting — including the multi-file features Composer generated in one shot that you never traced by hand.
This is a real advantage over recording your screen: a recording only captures the flows you personally remember to click. If Cursor's Agent added an admin panel, a webhook handler or a settings page you've never opened, a code read finds it and documents it; a screen recording misses it entirely. See user guide from code vs screen recording.
What three artifacts do you get for a Cursor app?
From one read of your Cursor project you get:
- End-user guide — plain-English help for your users, with share link and access code, in-app Help widget and portable export. $19/month ($12/month founding for the first 100).
- Founder runbook — how the app is structured and how to run and extend it, useful when Cursor's Agent scattered logic across files. Free.
- Code-verified fix-list — ranked, cited issues you can hand straight back to Cursor's Agent to fix, closing the loop in the same tool that built the app. Free.
An honest note
Because your code is already local, you can run Explain My Build on the folder entirely in your browser or via the CLI, so it never leaves your machine — a good fit for private Cursor projects. It's a code read, not a live penetration test, and you review the end-user guide before publishing it.
Frequently asked questions
How do I document an app I built with Cursor?
Point Explain My Build at your existing repo or local folder — Cursor already gave you real code on your machine. It reads the source (including the multi-file features Cursor's Agent built) and writes an end-user guide, a founder runbook and a code-verified fix-list. You review the guide before publishing.
Cursor's Agent built things I don't understand — can I still document them?
Yes — that's the point. Reading the code covers every feature the Agent and Composer generated across your files, including ones you never opened. You don't have to understand the code first; the tool explains it in plain English and cites where each claim comes from.
Is my private Cursor project safe to document?
Yes. Because your code is already local, you can run Explain My Build on the folder entirely in your browser or via the CLI, so the code never leaves your machine. Public repos read on our server are deleted, not stored.
Turn your app into a guide your users can follow.
Paste a public GitHub link and get your founder runbook + fix-list free in about a minute — no sign-up. Private code runs entirely in your browser.