What this tool does
Google Workspace keeps audit logs of every sign-in, OAuth consent, Gmail delivery, Drive action and admin change. After a phishing email, those logs answer the only question that matters: did someone get in, and what did they do?
Drop the exports from the Admin console (or the Reports API / GAM) and the analyzer runs detection rules across all of them: it correlates a risky sign-in with what followed — forwarding to an outside address, filters hiding invoices, a third-party app reading mail, mass downloads, admin changes — and gives an overall verdict, the evidence, a timeline and the remediation steps.
What it detects
- Sign-ins: Google's suspicious-login and leaked-password events, sign-ins from hosting / VPS networks (adversary-in-the-middle phishing), new countries, impossible travel, failed-login bursts, 2-step verification turned off, recovery email or phone changed.
- Mail: forwarding to external addresses, Gmail filters that forward or hide invoice / payment mail, delegates and send-as aliases, auto-forwarded messages, finance mail trashed in bulk, outgoing "new bank details" emails.
- OAuth: third-party apps granted Gmail, Drive or admin scopes (illicit consent grants), apps reading mail at volume, domain-wide delegation and app allow-listing.
- Drive: mass downloads, bursts of external sharing, public links, ownership transfers, Google Takeout exports.
- Admin: new super admins and admin roles, SSO / SAML changes, admin password resets, new users, mail routing and email monitors.
- Each finding carries its MITRE ATT&CK techniques. The rules are plain JSON (crates/gws-wasm/rules/rules.json in the source), so they can be reviewed and reused.
Limitations
- The verdict is only as good as the logs provided: a clean result on the sign-in log alone says nothing about Drive or OAuth. The tool lists the sources that were missing.
- Detections are heuristics. Travelling users, VPNs and legitimate apps cause false positives; a careful attacker can stay below the thresholds.
- Admin console CSV exports depend on the columns you selected and your console language: English column names and event titles are supported; the Reports API JSON is the most complete input.
- Country and ASN come from Google's network information: without them (some CSV exports), new-country and impossible-travel checks cannot run.
- Message content, Vault exports and the Security investigation tool's actions (deleting mail) are out of scope.
FAQ
Are my logs uploaded anywhere?
No. The analyzer is Rust compiled to WebAssembly and runs in a Web Worker in your browser. Files are read in chunks from your disk; there is no upload endpoint and no account.
How do I know if a Google Workspace account was compromised?
Look for a sign-in that doesn't fit (hosting network, new country, impossible travel) followed by persistence: forwarding to an outside address, a filter hiding invoice or payment mail, a new third-party app with Gmail access, or 2-step verification turned off. This tool checks those patterns and correlates them per account.
How do I find a Gmail forwarding rule set by an attacker?
The User log records "Out of domain email forwarding enabled" with the destination; the Gmail log shows messages auto-forwarded. Filters themselves are only visible in the user's Gmail settings: export them with the Gmail API or GAM and drop them here.
What is a suspicious OAuth app in Google Workspace?
An app a user authorised that can read or send their mail (https://mail.google.com/, gmail.readonly, gmail.settings…) or access all of Drive, especially right after a risky sign-in or with a lure name like "PDF viewer". Its token keeps working after a password reset until it is revoked.
Why can an attacker get past 2-step verification?
Adversary-in-the-middle phishing kits relay the real Google sign-in page: the victim types the password and approves the prompt, and the kit keeps the session. The sign-in then appears from the kit's server — usually a hosting provider — which this tool flags.
The verdict says compromised. What now?
Follow the remediation tab in order: contain the account (sign out sessions), revoke the app tokens, remove forwarding and filters, reset the password, warn the suppliers or customers who received fraudulent messages and stop pending payments. Then get incident response help to scope what was accessed.
Is this tool affiliated with Google?
No. It is an independent tool that reads the export formats Google documents. Google Workspace and Gmail are trademarks of Google LLC.