HETTIE ROWLEY · STUDIO OBSERVATORY

The AI Observatory

A live look at how search and AI-related crawlers reach this artist's website — and whether qualifying requests successfully retrieve its public information.

Connecting to Observatory…
ROLLING 24 HOURS MEASUREMENT CONTRACT 1.5
01 / LIVE MEASUREMENTS

Evidence from the website's Cloudflare edge · Updated approximately every 15 minutes

01 · IDENTITY

—

Corroborated crawler systems

Distinct autonomous crawler systems meeting the published-IP qualification rule.

02 · ACCESS

—

Qualifying machine accesses

Observed requests classified as credible autonomous machine activity.

03 · RETRIEVAL

—

Successful business retrievals

Qualifying GET requests returning 2xx for approved business-information pages.

04 · OUTCOME

—

Business retrieval success rate

Based only on qualifying scored attempts; no attempts means N/A.

02 / CORROBORATED ACTIVITY

Which systems visited?

Claims supported by a matching operator-published crawler IP range. This is network corroboration, not cryptographic proof of individual products.

Crawler Provider Accesses Business successes
Loading crawler measurements…
03 / DISCOVERABILITY

Can machines reach discovery files?

Separately scored requests for approved resources such as robots.txt, llms.txt and sitemaps.

—

Awaiting machine-discovery results…

A 2xx response counts as success; 4xx/5xx counts as failure. Redirects are neutral.

04 / OBSERVED SEPARATELY

Other machine activity

Recognisable requests that do not qualify for the autonomous headline figures — including uncorroborated identities and user-requested fetches.

— Diagnostic accesses
Reported crawler Classification Accesses
Loading diagnostic activity…
05 / HOW TO READ THIS EVIDENCE

Observed retrieval, not assumed influence.

01

Observe

A passive Cloudflare Worker measures selected crawler requests and origin responses without relying on browser tracking pixels.

02

Corroborate

A User-Agent label alone is insufficient. Headline reporting requires a match with the operator's published crawler IP ranges.

03

Qualify & score

Approved public pages and discovery files are evaluated separately. Successful retrieval does not prove AI understanding or recommendation.

Last evidence snapshot:

Built for Hettie Rowley using the Sydney Business Web AI Observatory. Synthetic tests are excluded; results cover only the displayed 24-hour window.