How it works
GuardScope processes only user-requested scans, combines multiple security signals, and returns a report designed for quick decisions.
GuardScope adds a focused security panel to Gmail so users can analyze suspicious messages without leaving the inbox.
The extension should request only the permissions needed for Gmail integration, local extension state, and calls to GuardScope services.
GuardScope detects the currently visible email and prepares the panel, but it does not submit email content until the user initiates a scan.
The scan can include subject, body, sender, recipients visible in the message, headers available to the extension, URLs, and attachment names.
When the user clicks analyze, GuardScope sends the needed message data to the backend for AI-assisted review and security checks.
Email content is used to produce the requested report and is not stored in GuardScope databases after the response is produced.
GuardScope returns a risk score, verdict, evidence, and suggested next steps so the user can decide whether to trust, verify, report, or avoid the email.
Results are advisory. High-risk actions should still be verified through an independent channel.
Analysis pipeline
Provider results may vary, so GuardScope combines signals and presents advisory findings instead of absolute guarantees.
Urgency, impersonation, pressure, BEC patterns
SPF, DKIM, DMARC, MX, sender alignment
Threat feeds and reputation checks
RDAP, domain age, registrar signals
Return-path and sender anomalies
Lookalikes, typosquatting, homograph clues
Risk scoring
The score helps users decide what to do next, while keeping the final decision with the user.
GuardScope is live on the Chrome Web Store for user-triggered Gmail threat analysis.
Add to Chrome