Blog editorial policy

This page explains how posts on the AgentProof blog are researched and kept accurate.

Every fact comes from a named source

Posts cite the source for each fact at the end of the post or next to the claim. We prefer primary sources:

• vendor product, pricing and documentation pages
• security advisories from the project or vendor
• the National Vulnerability Database, the CISA Known Exploited Vulnerabilities catalog and the GitHub Advisory Database
• researchers' own write-ups and published reports

Each CVE number links to its National Vulnerability Database entry. Figures we could not confirm from a primary source or reputable reporting are left out.

Competitor comparisons use public information

Comparisons describe what each company publishes about its own product: features, plans and prices. Prices are the published figures on the date of research. We say when a company does not publish prices, and when it does something better than AgentProof. Sponsored benchmarks are labelled as sponsored.

Versions and prices change

Each post shows the date it was published. Security fixes, version numbers and prices change after that date. Check the linked advisory or pricing page before you upgrade or buy.

Corrections are made in the post

When we find an error, we correct the post and show an updated date next to the original one.

Security guides are defensive

Framework guides explain how to find and fix weaknesses in your own software. They describe incidents at the level of what happened and how to prevent it. Only test systems you own or are authorized to test.