See PhishMind on a sample URL
Paste a URL below to walk through what PhishMind does to a reported case — extract, detonate, capture, agent-review. Verdicts here are scripted samples; the real pipeline runs against your API key.
Try a sample:
1
Submit a reported email
POST a
.eml
to /v1/cases
with type=user_report.
URLs are unwrapped from SafeLinks and Proofpoint redirectors before detonation.
2
Sandbox detonation
Each URL opens in fingerprint-chromium under tinyproxy. Documents render in an
E2B Desktop microVM. Screenshots, video, HAR, DOM, and rendered pages are
persisted in object storage.
3
Agent verdict + summary
An OpenAI Agents SDK loop reviews artifacts via tools, returns a verdict,
confidence, indicators, and a written summary. Cases auto-cluster into
campaigns when signals match.