The parts that get attacked are gone

There is no admin area

A content management system puts a login page on the public internet and asks it to hold the line against everybody who finds it. That page is the most attacked thing most businesses own. Here, content is edited through a private workflow and published as files, so there is nothing at your address for anyone to sign in to.

• No login page, no password reset, no session to steal.
• No plugin list, so no plugin to leave unpatched.
• No database answering questions from the public internet.

Updates are ours, continuously

Dependencies are watched and updated on the platform rather than on your schedule, and every update goes through the same checks a content change does. You are told what happened in the monthly record.

• Advisories are tracked against everything the site ships.
• An update that would break a page fails before it reaches you.
• Rolling back means rebuilding your current content against the previous platform version, so a rollback never discards an edit you made.

The browser is told what to enforce

A content security policy is built with the site and sent with every page, naming exactly what is allowed to run and where anything may be loaded from. Anything injected into a page afterwards carries no authorization and is refused by the browser itself.

• Script runs only from this origin and only where it was authorized when the page was built.
• Strict transport security, frame refusal, and referrer limits are response headers, verified from outside.
• The policy is checked on every change, so anything that would weaken it stops there.

What you can show

• A monthly record naming open advisories, updates applied, and certificate expiry.
• A header scan run from outside the network, dated.
• Answers to a customer security questionnaire, prepared for you.