ChecksProtocols
launchesgroupscrews
ledgerdocssign in
Sign in to watch coins, devs and crews and get alerts in Telegram.
Every check runs the same nine tests in the same order, against the chain and the listing. Nothing is scored on vibes, and nothing about a project's team, intent or future is assessed — only what can be read right now.
A coin can pass every test and still go to zero. The tests say what exists, not what it is worth or where it is going.
Nothing here assesses a team or an intention. A low score means the checks that would back a claim came back empty, which can also mean a coin is simply new.
The report lists every reading it used and the id is a hash of them, so running the same address against the same chain state gives the same id.
The newest pairs on Solana, each scored on the listing tests the moment it appears. Age is the one test a new launch cannot pass, and that is the point of the column.
| # | TICKER | SCORE | BAND | LIQUIDITY | 24H VOL | AGE | ADDRESS |
|---|
A group is a set of addresses watched together — a launchpad's output, a chain of redeploys, everything one Telegram channel posts. Watchtower scores the whole set and tells you when the shape of it changes.
A crew is the wallets that keep showing up together across launches. The page reads them from the chain rather than from anyone's word, which means it needs a crew on file to read.
Every check this browser has run, with its id and its score. It lives on this device only — nothing is sent anywhere and nothing is kept on a server.
What watchtower reads, how it scores it, and where it stops.
Paste a Solana mint address into the bar at the bottom of any page and press check. Watchtower asks a public RPC node for the mint account, its supply and its largest holders, and asks the listing for the pair, its liquidity, volume, age and links. Nine tests run against those readings and the report shows every one of them, passed or not, with the number behind it.
A test that cannot be read scores zero and says so rather than guessing — a node that times out looks the same as a missing answer, so re-run a check that came back thin. Listing data covers pairs the listing knows about, so a token with no pair has nothing to read. Holder concentration is measured over the largest accounts a node returns, which includes pool and program accounts, and the report names them rather than silently counting them as holders.