About Aback Audit
Aback Audit reads a public website from the outside, the way a crawler would, and reports what it finds. It is free, needs no account and returns a scored, machine readable report.
What it is for
Most people find out that their website has a problem the hard way: a search engine drops it, an assistant cannot read it, or a visitor leaves before the page appears. Aback Audit turns that into a list of concrete, ordered actions, each one with the evidence behind it.
How it is built
The engine runs 55 deterministic checks across 7 categories. A check either observes something or it does not: there is no model in the loop, no guess about what a crawler would think, and no score that cannot be reproduced from the evidence in the report. Checks that need a rendered browser are skipped with a reason when no browser is available, rather than being answered with a guess.
The service runs on Cloudflare. The audits are executed by a queue of workers, the reports live in D1, and the website is served from the edge.
What it does not do
- It does not rank sites, track keywords or read backlinks.
- It does not ask AI assistants whether they mention a site.
- It does not send anything to the audited site beyond ordinary GET and HEAD requests.
- It does not change anything. Every recommendation is a suggestion.
The family
Aback Audit is part of the Aback family of tools. Aback Tools is a library of utilities that run entirely in your browser, and the two services share a design language and a way of explaining things.