LODGEOSINT ORDER A REPORT
METHOD

Exactly how the sausage
gets made.

We publish our method because a verification product you can't audit is just someone's opinion. Here is the full pipeline, the quality controls, and the limits.

01 / THE ENGINE

Reports are produced by user-scanner, an open-source OSINT engine (MIT licence) that we run in library mode inside our own infrastructure — never as a black box. For each site on our curated list, the engine requests the public profile URL for your handle and classifies the response:

Only public, unauthenticated profile-existence checks are performed. No logins, no private data, no scraping of profile content beyond the public URL.

02 / THE CURATED LIST

We scan 60 sites (Tier A) — the full developer stack, major social platforms, marketplaces, gaming and finance profiles — not "2,700 sites" of noise. Every site on the list earned its place by returning correct verdicts in our test programme.

91
MODULES BENCHMARKED, OCT 5 2026
0
FALSE POSITIVES ACROSS 3 TEST HANDLES
18
OF THE 60 SITES HAND-VERIFIED

Benchmark, run on our server on 5 October 2026: 91 site modules tested against 3 handles: two with known real accounts (torvalds, kaifcodec) and one that exists nowhere. No module claimed the non-existent handle. The 60 sites below are the ones that answered reliably from our server.

High-value platforms behind aggressive bot walls — reddit, facebook, instagram, tiktok and similar — are deliberately excluded until each one passes individual verification. A smaller list that is right beats a big list that guesses.

03 / THE 60 SITES

Every report checks exactly these sites. A starred site (*) is one of the 18 whose check was hand-verified against known accounts in our Oct 2026 benchmark; hits there are labelled benchmark-verified in your report, hits elsewhere unverified. A site pulled by the weekly health-check is skipped until it passes again, and your report lists exactly which sites it checked.

04 / CONTINUOUS VERIFICATION

Sites change their defenses constantly, which is exactly how OSINT reports go stale. Every week, an automated health-check re-scans the full curated list against:

Flagged modules are excluded from all future reports automatically. Your report lists its scan date because it is a snapshot — true at the stamp, decaying after.

05 / EMAIL ORDERS

For a $19 order placed on an email address, the part before the @ is checked as a username across the same 60 sites. That is what every email report includes today.

Not live yet: a check of the address itself against public registration and password-recovery endpoints, using holehe (open source, GPL-3.0, by megadose — credited here per its licence). It is built but not switched on; until it is, email reports do not include it, and this page will say when that changes. When on, it runs as an isolated process inside our infrastructure; no account is created, no email is sent to the address, and no password is ever touched.

Where a site exposes masked recovery hints (a partially redacted recovery email or phone fragment), the report shows exactly what the endpoint revealed — no more, no less.

06 / WHAT A REPORT IS — AND ISN'T

A LodgeOSINT report tells you where an account exists under a handle. It does not, and cannot, tell you who operates that account. The same username on two platforms is frequently two different people.

Built for defensive use: vetting a marketplace seller, a date, a hire, a renter. Using it to stalk, harass, or discriminate is prohibited by our terms and may be unlawful — and our guardrails are built to catch that pattern.

06 / DATA HANDLING