Find out how your site reads to AI systems.
Kanso checks the signals AI systems can read on your public website, estimates how citable your pages look, and turns the gaps into tasks you can track. It does not ask ChatGPT, Gemini or Perplexity about you, and it says so.
Free to start · No credit card required. See how it works
At a glance
Being found now includes being understood by AI
More questions reach an AI assistant before a search page. Those systems can only use what they can reach and read.
The signals are invisible
llms.txt, robots.txt rules and structured data never show on the page. A site can look fine to people and be unclear to a machine.
Crawl rules block by accident
A robots.txt written for another purpose can disallow AI crawlers without anyone choosing that.
Advice has no evidence
General guidance says what matters. It does not say which of your own pages or files has the gap.
From guessing to a list you can work through
Kanso does not change how any AI system ranks or cites you. It makes your side of the question visible.
- Wondering whether AI crawlers can reach your site
- Opening robots.txt and llms.txt by hand, when you remember
- No record of what you checked or changed
- Nine named AI crawlers checked against your robots.txt
- A readable status for llms.txt, structured data, headings and your sitemap
- Recommendations with their evidence, and tasks you can track to done
Two agents, two kinds of evidence
The GEO Agent looks at what AI systems can read. The SEO Agent covers the wider health of the same site.
GEO Agent
Checks the signals AI systems can read on up to 25 public pages, and estimates how citable they look.
- It produces
- A readiness analysis, a citation estimate and recommendations.
- Where it fits
- Inspect, estimate, recommend.
SEO Agent
Runs the public site analysis the readiness signals come from, and writes its own search recommendations.
- It produces
- Findings on titles, descriptions, canonical links, links and speed.
- Where it fits
- Alongside the GEO findings, for search fixes.
From your public site to a tracked task
Five steps, in the order the product runs them.
Inspect your public site
Kanso reads up to 25 public pages, plus robots.txt, sitemaps and llms.txt. It checks nine named AI crawlers against your rules and looks for structured data, titles, headings, canonical links and visible text.
Estimate citation potential
A model judges whether an AI assistant would cite your site for up to five queries inferred from your website, with a one-sentence reason each. It is Kanso's estimate, labelled as one.
Recommend, with evidence
Findings become recommendations showing what was observed, what was derived and what the data cannot establish. The wording is labelled as an AI-generated suggestion.
- 04You
Decide and act
You read, question or dismiss each one. A recommendation you pursue becomes a task that you or your team move from open to complete, with a record of who did what.
- 05Kanso, then you
Run the analysis again
Re-run it to see whether each signal now passes. A pass means the file or markup is there, not that any AI system uses it.
What you receive
Everything lands in your Kanso workspace, where you can read it before you do anything.
A readiness analysis
Pass, warning, missing or not measured for each signal, from llms.txt and AI crawler access to structured data and HTTPS.
A citation estimate
For a few queries inferred from your site, whether a model thinks an AI assistant would cite you, and why.
Evidence-backed recommendations
Each shows the observed data behind it and the limits of that data.
Action tasks
A tracked task with an audit trail of who changed its status.
Kanso reads and recommends. You change things.
Nothing in this workflow edits your website.
Kanso does
- Reads your public pages and the rules you have published
- Estimates citation potential and labels it
- Records the tasks you open
You decide
- Which recommendations to act on
- What to change on your site, and when
- Whether a task is complete
Kanso never
- Edits your site or opens a pull request for a GEO change
- Calls a GEO task verified
- Asks any AI product about you
Where this fits
Situations, not results. What Kanso shows depends on your own site.
You are launching a new site
Check whether AI crawlers are allowed in and whether llms.txt exists before you send people to it.
You think robots.txt is too strict
See whether it disallows each of nine named AI crawlers from your homepage.
You want a page to be quotable
Read the estimate and its reasons to decide which page to make clearer first.
How Kanso approaches AI search
Each point is something the product does today.
Observed, estimated and not measured stay apart
Readiness signals are deterministic checks. The citation view is a model judgment. Live AI answers are not measured.
Recommendations show their evidence
Kanso resolves the facts behind each one from its own data, and marks them observed, derived or a limitation.
It sits beside your SEO work
The same site analysis feeds the SEO Agent, so AI readability and search health are not separate projects.
What Kanso does not do here
AI search is easy to oversell. These are the limits.
It does not ask AI products about you
Kanso does not query ChatGPT, Gemini, Perplexity or Claude, or show what they say about your brand.
It does not track citations over time
There is no history of rankings, citations or share of answers. Re-running the analysis shows your signals, not your standing.
It does not control how AI ranks you
No tool can. Kanso checks what your site exposes. Whether a system cites you is outside its reach.
The estimate is not a measurement
It is Kanso's own model judgment for queries it inferred from your site, not a result from any AI product.
It cannot verify a GEO task
Formal verification covers page-level SEO conditions today, so a GEO task is never marked verified. Run the analysis again instead.
Related use cases
Turn Search Console Data Into Content
Turn Search Console opportunities into long-form article drafts in your brand voice, for you to review and publish.
If the gap is missing content, the Writer Agent drafts articles from your Search Console queries.
Ship Technical SEO Fixes
Turn supported technical SEO findings into GitHub pull requests you review. Kanso never merges or deploys.
For supported title, description and canonical fixes, the Coding Agent prepares a pull request for you to review.
Questions about this use case
What does Kanso check for AI search?
Up to 25 public pages for llms.txt, the rules for nine named AI crawlers, whether the site can be crawled, structured data, titles, headings, canonical links, a sitemap and HTTPS. Separately, the GEO Agent estimates citation potential.
Can Kanso tell me whether an AI system cites me?
No. Kanso does not measure live AI answers or track citations over time. It checks what your site exposes and estimates how citable your pages look.
Does a passing llms.txt check mean AI systems use it?
No. It means the file exists and is readable. Kanso does not claim any AI system reads it.
Does Kanso change my website?
No. A recommendation can become a task that you or your team carry out. Kanso does not edit your site or open pull requests for GEO changes.
Do I need Search Console?
Not for the readiness checks, which read your public site. Connecting it, with read-only access, gives the SEO Agent more to work from.
How often does the analysis run?
The public site analysis is refreshed weekly and whenever you run it. For the citation estimate: Runs when you run the GEO Agent, and a daily run is built in but currently paused. It is Kanso's estimate, not a result from any AI product.
See plans and pricing · All use cases
See how your site reads to AI systems.
Enter your website and Kanso checks the signals AI systems can read, then lists what is worth fixing.
Get started free · No credit card required