Cloudflare lab / synthetic content / not financial advice
Content
Observatory
FIELD EDITION
CONTENT & CONTROL

THE READING ROOM / AN ACCESS EXPERIMENT

Open to readers.
Intentional about reuse.

Good publishing makes ideas accessible. Clear boundaries make their reuse a deliberate choice. Explore the difference between a stated preference and an enforced decision.

Observe a real request

From the library

Three pages. Three access stories.

Live access checks

One click. One same-origin GET.

REQUEST DESK

Follow the request,
not the assumption.

Start with a visitor, read the preference, then send a synthetic training marker to the same content.

Visitor and robots checks send no custom header. The other checks send X-Demo-Crawler: training. Requests omit credentials, bypass cache and time out after 10 seconds. No checks run automatically.

Ready. Choose a check to begin.

RESPONSE INSPECTORNo request yet

The console is empty.

Results appear only after a real request. No statistics are generated here.

Request / correlation
Not sent
Elapsed / full body
Not measured
Cloudflare Ray ID
Not received
X-Demo-Origin
Not received
cf-mitigated
Not received
Response preview / plain text / up to 4,096 characters

What this proves. What it does not.

Synthetic request marker, not verified crawler identity. This demonstrates WAF enforcement, not the AI Crawl Control dashboard toggle. Analytics are in TME; this page does not fabricate statistics.

robots.txt signals preferences, not enforcement. With the demo WAF rules disabled, marked requests reach content. This Worker never generates HTTP 403 or 402.

HTTP 402 here is a synthetic licensing notice, not Pay per Crawl. No payment is requested, processed or collected. TME analytics are demonstration-environment analytics, not this page's traffic statistics.