Score your AEO structure before your site ships.
ailk audit runs the same schema and structured-data checks in your terminal and in CI. Free, in the Apache 2.0 foundation. Catches a missing schema or an unscored page before it costs a citation.
The page is the unit of citation, not the site.
When an AI system recommends a business, it read pages, attributed them to a brand, and quoted what it trusted enough to cite. A site can rank well, load fast, and still never get quoted, because the page itself didn't carry the structure a model needs to identify the brand behind it and trust the claim. That structure is checkable. Missing schema. Ambiguous entity markup. A page that never emits JSON-LD. These are AEO problems, and AEO problems are silent — the site doesn't crash, it just doesn't get cited, and nothing tells you why. ailk audit exists to make that failure loud instead of silent. It runs against the typed schema registry every AILK page ships with, checking the structural properties a model relies on to extract and attribute a claim. It reports the gaps before the site goes live, not three months after an agency notices the client stopped showing up in AI answers.
ailk audit checks the structural properties AI systems need to cite a page.
It runs against AILK's typed schema registry of 60 page-type entries, scoring each page for the specific things a model looks for before it trusts a claim enough to quote it.
Schema presence and validity
Confirms each page type carries the right JSON-LD, validated per-type against the typed registry — not a generic schema stub.
Entity clarity
Flags a page where the brand behind the claim isn't unambiguous, the exact gap that leaves a model unable to attribute what it's reading.
JSON-LD emission
Verifies the page actually emits the markup at build time, not just that the frontmatter intends to.
FAQ and glossary structure
Checks that FAQ and glossary content is marked up so it's extractable as a direct answer, not buried in undifferentiated prose.
What a failed check actually looks like
Four failure patterns ailk audit is built to catch — the ones that cost a citation without ever throwing an error.
The missing schema
A page ships without the JSON-LD its page type requires. It renders fine for a human visitor and stays invisible to a model deciding what to cite.
The ambiguous entity
A page never clearly states which brand it belongs to. A model reading it can't attribute the claim, so it looks elsewhere for a source it can name.
The silent JSON-LD failure
The frontmatter is correct, but the build step never emits the markup. Nothing breaks visibly; the page just quietly stops being structured data.
The unscored page
A page gets added after the last audit runs and never gets checked. In CI, this is the exact gap the parity and schema checks catch automatically.
Pre-launch and free, instead of post-hoc and paid
Enterprise AEO platforms charge $295 or more a month to measure whether a site is already being cited, after the structural gap has already cost the citation. ailk audit measures the structure itself, before the site ships.
Cost
When it runs
What it measures
License
The two aren't the same tool. Post-hoc citation tracking still has value once a site is live. ailk audit's job is making sure the structure is right before that tracking ever starts.
Structural readiness isn't a premium feature.
AILK's OSS layer is Apache 2.0 — not source-available, not free-until-you-make-money. It isn't deliberately limited so a paid tier can earn. Paid tiers on AILK add capability. They don't gate access, and a structural AEO check is access, not capability. That's why ailk audit ships in the free foundation and runs in CI on every change, not behind a login. A developer or an AEO-practitioner agency should be able to verify the structure is right without a subscription standing between them and the answer.
Common questions about ailk audit
Is ailk audit different from a schema markup validator?
Yes. A schema markup validator checks whether existing JSON-LD is syntactically correct. ailk audit checks something upstream of that: whether the right schema exists at all for a given page type, whether it's emitted at build time, and whether the entity behind the page is unambiguous. It's a structural readiness check, not a syntax check.
Does ailk audit replace a structured data testing tool?
It complements one. A structured data testing tool validates a single page's markup after the fact. ailk audit runs across the whole site, pre-launch and in CI, against AILK's typed schema registry, so gaps get caught before a page ships rather than found one URL at a time afterward.
What does 'AEO scoring' mean in ailk audit's output?
The score reflects structural readiness for AI-mediated discovery: schema presence, entity clarity, and JSON-LD emission per page type. It's not citation volume or ranking position. It answers whether the page is built so a model can read and attribute it, not how often the page is already being cited.
Can I run an AI search audit against a site I don't own?
ailk audit is built to run against any AILK site, including this one, since it checks structure that's already public in the rendered page and its JSON-LD. Running it against a non-AILK site isn't the intended use case; the checks are built against AILK's typed schema registry, not a generic crawler.
Clone the OSS and run your first audit.
ailk audit ships in the Apache 2.0 foundation. Clone the repo, scaffold a site, and run the check before you ship the first page, not after you notice you're invisible.