Load Email File
ML Model Check
The second verdict. The classifier chosen above reads the email TEXT and states its own verdict here when you click “Run ML model check”, next to the parser’s verdict and with the model’s own confidence. The Scan Verdict below stays the parser’s — the two are shown side by side, never merged.
Scan Verdict — the parser’s verdict
No File
0%
{}
Waiting for a file… drop one on the page or pick one above.
Settings — brands and blocklists
Extend the built-in lists with your own data. JSON arrays are used
(they persist in this browser only). To import the UrlGov “top 1,000 most phished brands”
CSV: convert its rows to [{"name":"…","domains":["…"]}] and paste it in
— or set an update URL below that returns {"knownBad":[...]}.
Parsed Email Sandbox
🔒 Isolated Preview
From:
None
Subject:
None
Message Content:
Extracted Links Ledger (Clicking Disabled):
- No links extracted yet.