How Centinel evaluates automated traffic
Your integration sends request evidence to Centinel. Centinel evaluates that evidence against your policy, then your CDN, reverse proxy, or application applies the returned response.
Request a site auditClear rules for crawler traffic.
You decide which bots can reach your site, page by page. Your CDN or WAF stays where it is.
Scrapers retooled. So did we.
Our background in building crawlers informs how we investigate browser automation. We test tools and review their behavior as browsers, proxies, and evasion methods change.
How scrapers retoolNo single check catches a modern scraper
Useful detection combines evidence with the resource being protected. Your evaluation should test both the automation decision and the response a legitimate user receives.
Evidence from several layers
Request context, crawler identity, and available browser and session signals contribute to the decision. No individual fingerprint establishes whether a request is welcome.
Your access policy
Your team decides what each resource should allow. The integration applies the returned response, while your operators retain responsibility for policy and review.
User impact you can test
Include ordinary browsing, accessibility tools, and approved automation in your evaluation. Check challenge completion, latency, and unexpected blocks before widening enforcement.
One call, from the stack you already run
The integration determines where enforcement happens. An edge deployment can respond before origin; an application deployment makes the decision inside your server. Test the chosen path, including its timeout and error behavior.
Where Centinel plugs into your stackYour edge
Sends the request it already has
A supported CDN, reverse proxy, or application integration sends the protected request's context to Centinel. Available connection evidence depends on that integration and any upstream proxy.
Centinel
Decides
Centinel checks the request against policy and available automation evidence. Browser and session signals contribute where collection is supported and has occurred; direct API clients do not supply a browser payload.
Your edge
Acts on the answer
Serve the page, send the visitor to a check, or refuse. Centinel only returns the answer. Your own infrastructure is what talks to the visitor.
Which evidence contributes to a decision?
Browser evidence
Supported browser integrations collect runtime and session signals alongside request context.
This adds evidence about clients that present browser-like headers; coverage depends on where collection runs.
Read more: Browser evidenceClient validation
Client-reported properties are evaluated for tampering and inconsistencies with other observations.
A plausible browser claim can be checked against independent evidence instead of being accepted on its own.
Read more: Client validationProof of execution
The checks run inside an obfuscated virtual machine whose code paths change on every build.
Execution evidence adds another condition to evaluate; it does not by itself establish who operates the browser.
Read more: Proof of executionCrawler identity
Declared crawler identities can be checked against operator-published network or DNS evidence.
Your policy can distinguish verified operators from clients that only copy their names.
Read more: Crawler identity
Detection review
Request and decision evidence supports investigation of suspected bypasses and unexpected blocks.
Review the affected workflow and legitimate-user impact before changing a control.
Threat intelligence
We track the scraping tools and the communities that build them, because we used to be in them.
Changes in those tools inform what needs testing and which detection assumptions need review.
Read more: Threat intelligence
Protect the pages scraping costs you most
Usually that is paid content and the pages your customers sign in to.
- Paid content and downloads
- Account and signup flows
- Search and archive pages
- Comments and community areas
- Checkout and subscription paths
Want to see what reaches your site?
A free audit tests which scraping tools can retrieve pages on your domain, then emails the reviewed findings with fixes to consider.