MIRAGE
Manufactured Influence Recognition & Amplification-Graph Evidence
Tell a real crowd from a rented one. MIRAGE maps the network behind an account’s engagement and scores it for coordination — in two directions.
Just want to vet one account? Check a single account → Ready to map a whole crowd around a target? Pick a mode below. New to the idea? How it works →
Coordinated boosting — FOR a target
A rented crowd inflating an account: synchronized likes, reposts, and templated praise that fake momentum.
Open this mode →Coordinated attacking — AGAINST a target
A pile-on manufactured against an account. MIRAGE flags the coordination in the behavior — never the opinions, and never to dismiss genuine criticism.
Open this mode →A rented crowd distorts what real people see.
Bought engagement — for a candidate, or against a critic — hijacks the algorithm and warps the conversation. MIRAGE gives you the evidence to see it clearly and act responsibly.
MIRAGE only documents coordination — it can never run, simulate, or hide one. Detection and evidence only.
Every Discern Report is deterministic and SHA-256 signed — a third party can re-run it and get the same result.
Runs entirely in your browser. No account, no uploads, public data only. Your case never leaves your device unless you share it.
- ▪ ≥3 engagements before an account counts as part of a ring
- ▪ elevated only when ≥3 of 5 signal families converge
- ▪ scores only what you paste — no auto-collection, no scraping
- ▪ public data only; one-time engagers are never enumerated
- ▪ always "indicators of coordination" — never "bot," "fraud," or "proven"
Detection over operation · Evidence over accusation · Document over speculate