How Qlinnt checks websites
A finding needs evidence. A pass needs a completed check.
Qlinnt reviews the public behavior of a website. This page explains how to read its results, where the scan stops and how the implementation is tested.
Published by Qlinnt. Product and editorial contact: hello@denilson.me.
Reviewed
Where the inspection boundary sits

- Public evidence Qlinnt can inspect
- The submitted URL and up to four priority internal pages on the same origin, plus supported public configuration checks. Each report states which checks completed.
- Private or manual review still needed
- Source code, database policies, authenticated flows and business logic require separate review. A public scan cannot certify security or guarantee search rankings.
What the scanner can observe
The scanner starts at your public URL and may inspect up to four priority routes on the same origin. The report records pages, attempted modules and checks that did not reach a conclusion. Coverage varies by route and dependency.
It does not sign in, review your repository or inspect database policies. The read-only browser blocks state-changing requests; this can limit what a page loads. Some security checks inspect public resources without submitting forms.
Read outcomes before drawing conclusions
Found means the check produced an observation. Passed means that particular check completed without its finding on the measured page. Inconclusive, blocked and error are not passes. Not applicable means the check’s condition did not apply.
Confirmed observations, potential issues and informational results have different meanings. A high fraction of passed checks does not cancel a critical finding or establish that a website is secure.
How a correction is verified
A verification is a new measurement of the same URL. Comparison groups results by check and page. A previous finding is only marked resolved when that check explicitly passes on that page; missing, blocked or unmeasured pages remain unverified.
How this behavior is tested
The repository includes deterministic regression fixtures, controlled browser tests and report access tests. The browser integration suite supplies its own HTML and network responses, so it can repeat the same scenarios without scanning a customer site.
A local fixture suite is not an estimate of real-world accuracy. No percentage of overall security or infallibility follows from these tests.
npm test -- worker/src/scanner.integration.test.js src/lib/scan-comparison.test.ts
npm run evaluate:scannerShared reports and responsible use
Report URLs are shared links: anyone holding a link can open the available report tier. Exclusion from search does not provide authentication. Keep links private if you do not intend to share them.
Only assess sites you own or have permission to test. Do not paste exposed credentials into prompts or public discussions. Use the contact below to report a reproducible issue, with secrets removed.
