Getting started

First-run onboarding

When ShieldWays installs, it opens a four-step onboarding page in a new tab. You can skip at any point — protection works with the defaults.

1. Welcome

   A quick introduction to what ShieldWays does, with a plain-language data statement you explicitly agree to before continuing: signed out, nothing ever leaves your device; if you sign in, we sync your settings and receive content-free detection events — type, action, domain, time. Never your text.

2. Pin it

   Pin ShieldWays to the toolbar so your protection status is always visible: click the puzzle-piece icon, then the pin next to ShieldWays.

3. Test drive

   Try the detection engine on sample sensitive text and watch it get caught — the same scan that runs on every AI site you visit.

Signing in

Click the ShieldWays icon in your toolbar and choose Sign in. This opens the ShieldWays dashboard in a normal browser tab, where you sign in with your account (or create one). The extension detects the dashboard session automatically — there’s no separate extension login, no copied tokens.

Signing in unlocks syncing:

• Your AI Guard settings (categories, actions, per-tool rules, custom keywords) follow you across devices.
• Content-free detection events appear in your dashboard audit log.
• Your plan tier (Free or Pro) is applied to the extension.

Connecting to the dashboard

The dashboard is where you configure ShieldWays in depth. Changes you save there — enabled categories, actions, per-tool overrides, custom keywords and regex — sync to the extension automatically. The extension also keeps a local copy of your configuration, so protection keeps running with your last-known settings even when you’re offline.

The pill

On supported AI sites, ShieldWays shows a small floating Protected pill. It’s rendered in an isolated shadow DOM, so the page’s own styles and scripts can’t interfere with it.

• Green dot — AI Guard is active and watching this page. Nothing sensitive has been flagged.
• Amber, pulsing dot — something in your current input was flagged. The pill shows a count of detections.

The popover

When AI Guard flags your input — on paste, on pressing Enter, or on clicking the site’s send button — it intercepts the send and opens a popover explaining exactly what it found, by category. Depending on your configured action you can:

• Send anyway — you reviewed it and it’s fine. Your choice is logged (content-free) in the audit trail.
• Redact & send — replace the sensitive values with placeholders, then send. Available on the Pro plan.
• Edit first — close the popover and change the text yourself before sending.

Next, learn how detection actually works in AI Guard.