How we check airdrop geo

A venue's restricted list is usually undated, so the only way to know it moved is to have taken the previous reading yourself. A record like that cannot be assembled after the fact, which is why the method matters more than the page count.

Where this site actually stands today

Stated first, because a methodology page that describes an intention in the present tense is the failure it is supposed to prevent. Updated 30 September 2026.

Six of the nine board venues are machine-polled hourly (extended, lighter, ondo, ostium, pacifica, polymarket); three stay hand-authored because their pages are JS shells or empty ToS. The ledger opened on 30 August 2026.
That first reading is a baseline, not a change. It was filed for review and paused the offer until a person authored the mask — which is now live.
Scheduled re-reading runs hourly via airdropradar-probe.timer on this host. Each offer page still shows how many readings stand behind it.
No page here carries a machine-verified freshness date, because no page is yet projected from an approved row. A date on a page is either a reading we took or nothing.

What the monitor does on every check

The same four steps, on the interval configured for each source. This describes the built system, not a plan.

Fetch the venue's own page, never a summary of it.
Extract the country set and hash it separately from the page body.
Compare against the last reading we trust, not merely the last one.
Stop and ask a person whenever the set moves or we read the page worse than before.

What we refuse to do

Each of these would make the record look better and mean less.

Move a freshness date without re-verifying the fact behind it.
Map a restriction on part of a country onto the whole country.
Invent airdrop listings. A country page exists only for a venue we have actually read, and only restates that reading.

How an offer is scored

Published because no competitor publishes theirs. A score with no visible rubric is an opinion wearing a number, and the reader has no way to disagree with it. Nothing here is weighted by what a venue pays.

Reading evidence	0-30	Does the venue publish a restricted list at all, and does it date it? An undated list scores lower because it can move without notice.
Our own observation	0-25	How long we have been snapshotting it, and how many changes we have caught first-hand. Zero on day one, by construction.
Payment reality	0-20	Whether the referral programme pays cash, on what cycle, and whether it is open or invite-only. Points-only scores low here and is said so plainly.
Access friction	0-15	KYC, minimum volume, and whether a referral code is tied to a wallet before first use.
Operational risk	0-10	Trading halts, withdrawal incidents, and unexplained changes in published terms.