LIVE — ROLLING 24-HOUR EVIDENCE
AI Observatory - Who Has Actually Been Here?
This Observatory records qualifying search and AI-related machine retrieval activity observed against the Djarindjin website. The figures below show a filtered evidence view — not every automated request that reached the server.
Distinct autonomous crawler systems whose identities met the public corroboration threshold during this rolling 24-hour period.
Successful qualifying GET retrievals of known Djarindjin HTML pages and substantive PDF documents.
The proportion of qualifying business-information retrieval attempts that returned a successful response.
Successful retrievals of deliberate machine-facing discovery resources such as robots.txt and recognised sitemap XML files.
The totals above are measured observations, not estimates. The table below shows the individual machine systems that currently satisfy the public evidence threshold.
MACHINES OBSERVED
Which Search & AI Crawlers Produced the Evidence?
Every qualifying autonomous system in the current rolling 24-hour window is shown below. The list changes as real crawler activity enters and leaves the window.
| Machine | Role | Autonomous Accesses |
Business Retrievals |
Success Rate |
Discovery Retrievals |
|---|---|---|---|---|---|
| Loading current machine evidence… | |||||
A system appears when qualifying activity from it is observed inside the rolling 24-hour window and disappears again when that activity falls outside the window.
That is deliberate. A larger number is easy to produce; a defensible number is more useful.
Synthetic requests used to test the Observatory are excluded from public evidence.
User-triggered fetchers and diagnostic tools are excluded from the autonomous headline metrics.
A crawler name in a User-Agent is not enough. Off-edge headline evidence requires independent provider-network corroboration.
Assets and unclassified URLs do not inflate business-information retrieval figures. Unknown business resources fail closed.
The Observatory measures observable retrieval. It does not by itself prove that an AI system understood, indexed, stored, cited, recommended or endorsed the retrieved information.