
Summary
This rule detects inbound emails that impersonate legitimate invoices or billing notices and direct recipients to a web domain via a crafted hyperlink to harvest webmail credentials. It analyzes the email body for links and decodes the URL query parameters, looking for a pattern where _task equals mail and _action equals get, and where _mbox and _token are present. The destination domain (href_url.domain.root_domain) must differ from the sender's domain root and must not be part of the organization’s approved domains. If these conditions are met, the rule flags a credential-phishing attempt that uses social engineering and an out-of-band pivot to a credential collection page. Detection methods include URL analysis (to confirm the phishing-like query parameters and redirection), sender analysis (to identify domain mismatch), and content analysis (to inspect the link structure within the message). This rule is effective for catching webmail credential harvesting campaigns embedded in inbound messages but may require suppression/validation in cases of legitimate cross-domain integrations; supplementary checks (e.g., DMARC/SPF alignment, sender reputation) can help reduce false positives.
Categories
- Web
Data Sources
- Web Credential
Created: 2026-11-18