Pass the AI audit
Short answerThe automated audit checks security, structure, documentation, dependencies and whether your app builds. You need at least 75/100 with no failing check. Fix the findings it lists, upload again, and ask Remy if a finding isn't clear.
The audit is the first gate every project goes through. It's automatic, fast, and it tells you exactly what to fix.
What the audit checks
- Security — secrets in code, unsafe patterns, risky settings.
- Structure — how the project is organised.
- Documentation — README and setup instructions.
- Dependencies — outdated or vulnerable packages.
- Build — whether the app actually builds.
What you need to pass
- A score of at least 75 out of 100, and
- No failing check. A single failing check stops the project, even with a high score.
Warnings don't stop you — but buyers see them on your listing, so fixing them makes your app look stronger.
The most common findings — and quick fixes
- Secrets in the code. Move them to environment variables and commit an example file (`.env.example`) instead.
- Missing or thin README. Write clean-machine setup steps: install, configure, run.
- Outdated packages. Update them and test.
- Build errors. Make sure a fresh install builds without your local files.
- Leftover test or debug code. Remove it.
Fix and run it again
Fix the findings, upload the new version, and the audit runs again. There's a daily limit on how many audits you can run, so fix everything you can before re-uploading.
💡 Don't understand a finding? Ask Remy in the chat — paste the finding and he'll explain what it means and how to fix it.
Why the audit helps you
A good score is shown on your listing. Buyers trust audited code — and trust sells.
Questions
Can I pass with warnings?
Yes. Warnings don't stop you, but buyers can see them, so fixing them helps sales.
How many times can I run the audit?
As many as you need, within a daily limit per builder. Fix everything you can before each upload.