Every check a Deep Audit can run is listed in a public catalog at seoscoreapi.com/checks. There are 14,357 of them today. You can read what each one tests, what counts as a pass, how much it weighs in the score, and whether it is automated or scored by the AI analysis pass. You don't need an account to read any of it.
What is in the catalog
The catalog is organised the same way the score is: nine sections, each split into subsections, each subsection holding its individual checks.
| Status | Checks | What it means |
|---|---|---|
| Automated | 8,247 | A deterministic checker in the engine tests it |
| AI-assisted | 6,080 | Scored in the Deep Audit's AI analysis pass |
| Planned | 30 | Listed, not yet scored |
The planned checks are shown as planned. A check that isn't live yet doesn't count toward anyone's score, and the catalog says so instead of leaving it out.
The same data is available as JSON for anyone who would rather read it in code: catalog totals and the full catalog.
Why publish it
A score you can't inspect is a number you have to take on trust. "Your site scored 62" is only useful if you can find out what was measured. When the list is public, a low score points at specific things, and a high score can be checked the same way.
It keeps our claims honest. "14,000+ checks" is easy to write on a pricing page. Publishing the list means the number has to be real, and anyone can count.
Developers need it. If you put an audit in a CI pipeline or a client report, you need to know what a failing check means before it blocks a deploy at 5pm. Reading the definition beats guessing from a label.
Agencies get asked. Clients ask why their score dropped. Being able to link to the exact check, with its pass condition, ends that conversation quickly.
What we gave up
A competitor can read the list. We decided that doesn't matter much. A list of checks is not an engine: the work is in fetching pages reliably, rendering them, scoring thousands of results consistently, and keeping all of it running. Hiding the list would protect very little and would cost every customer the ability to verify what they're paying for.
How to use it before you buy anything
- Open the catalog and pick the section you care about, such as accessibility or structured data.
- Read a few checks in a subsection. Each one shows what passes and how it is run.
- Run a free audit on your own site and compare the findings against the definitions.
If the checks match what you'd want tested, a single Deep Audit is $12 with no subscription, and it is included on the Pro and Ultra plans.
Frequently asked questions
Does a Deep Audit run all 14,357 checks on every page?
No. It scores every check that applies to the page. A check about product schema, for example, doesn't apply to a page with no products. A typical audit scores several thousand, and the response reports how many were scored, passed, warned and failed.
Are the standard audit's checks public too?
Yes. The standard /audit endpoint runs 83 checks, and each one appears by name in the response with its result.
How often does the catalog change?
It is rebuilt from the engine whenever checks are added or changed, so the public pages match what is actually running.