Hall 01 · Night audit

Audited
while you sleep.

One command before bed. 11 agents sweep your whole codebase, fix what is safe to fix, and flag the rest.

89€one time · incl. VAT

Download link by email, seconds after checkout. Personal license, lifetime updates.

FULLMOON on a laptop: the product artwork, a full moon over a ridge with scan windows open across the sky

No. 04 · FULLMOON · The night audit

Hall 02 · What it does

What one night covers.

One night, one command, every category.

  • Security to SEO

    Security, accessibility, performance, dependencies, SEO, design, content and more. Run all of them or one.

  • 11 parallel agents

    Every category runs at the same time, so a ten-repo workspace is one wait instead of ten.

  • 25 safe auto-fixes

    console.log removal, colour tokens, em dashes, missing alt text, var to const. Never architecture, auth or pricing.

  • Verified before you wake

    Every fix is rebuilt, retested and regression-checked. Anything that breaks is rolled back.

  • A living report

    One HTML page with the findings, the fix history, and how this run compares to the last.

  • Any stack

    JavaScript, TypeScript, Rust, Python, Go and static sites.

Hall 03 · Living report

The FULLMOON living report on a laptop, close on the S3 Medium section with 4 remaining, the findings table listing globals.css across 3 repos and remotion/Root.tsx line 57

Real report · The findings table · 23 of 34 fixed

04aThe report

The report remembers the last one.

Every run writes one HTML page. It compares this run to the previous ones and shows what got better and what came back.

  • Audit journey

    Findings compared across runs.

  • Regressions

    Anything fixed once and back again gets flagged.

Hall 04 · Questions

Before you buy.

Something missing? Email help@raxxo.shop

What do I need to run FULLMOON?

Claude Code, CLI or desktop app. FULLMOON is a skill that lives inside your existing setup. Install it, type /fullmoon, done.

How long does an audit take?

A single repo is a few minutes. A whole workspace takes longer, but the 11 agents run at the same time, so it is one wait, not one per repo. You start it before bed and read the report in the morning.

Will it break my code?

No. Auto-fixes only touch mechanical patterns: console.log calls, colour tokens, em dashes, missing alt text, var to const. Architecture, authentication, pricing and business logic are never touched, they get flagged for you. Every fix is rebuilt and regression-checked before it stays.

Does it send my code anywhere?

Not to me. FULLMOON has no analytics and no phone-home. It runs inside your Claude Code session, so your code goes to the Claude API exactly as it does in any other Claude Code session, and nowhere else.

Can I run it on one project?

Yes. /fullmoon project my-app scopes it to one repo. /fullmoon category 3 runs a single category, security in that example.