Lawful cyber-forensic attribution & OSINT engine
AEGNYX‑TRACE ingests sealed evidence and produces an auditable Trace‑Map, a calibrated verdict, ranked attribution leads, competing hypotheses, and an explicit false‑flag assessment — built for authorized analysts, designed to be explainable in review.
Live Trace-Map console
Select a case to load its real engine output: Trace-Map, timeline, verdict, calibrated confidence, ranked OSINT leads, alternative hypotheses and false-flag risk.
Capability matrix
Module domains measured directly from the installed package at build time — network, host, OSINT, forensics, correlation, attribution and reporting.
Architecture
A deterministic, offline-by-default pipeline. Every derived fact carries an evidence hash from ingest to Trace-Map, so each conclusion is explainable and re-runnable.
Methodology & honesty
Each case resolves to one calibrated verdict with an explainable provenance graph. The engine states insufficiency plainly rather than over-claiming.
Reported accuracy and confidence are conformance results on generated ground truth — a SYNTHETIC-range target. They are not, and must not be read as, measured real-world attribution accuracy.
Attribution output is a lead for authorized human review. Consistency with an actor profile never excludes third-party imitation; every lead ships with its false-flag check and alternative hypotheses.
The default pipeline performs no live collection. Network collectors are recorded as skipped, never failed or fabricated; passive reasoning runs on sealed evidence.
Operates only on synthetic Forensic-Range data or on evidence the operator is authorized to analyze. No intrusion, no hack-back — passive and lawful by construction.
Dossiers are DRAFTs pending authorized analyst signature; admissibility is reserved to independent legal review and is never self-certified by the platform.