Tetrascore

LOOKUP.EXE

Token address

Token Wallet Dossier

Free and instant — a lookup against an index, not a call out to a chain.

Verdict

Paste a token address above.

What ran

A check that could not run says so here. “Not run” and “found nothing” are different answers and this never merges them.

CREATOR.DAT

Who made this

-

Tokens launched-

Reached a pool-

Enough to trade against-

Liquidity later pulled-

Their rate-

This chain's rate-

Read

-

A rate on its own is unreadable — 5% could be excellent or catastrophic. The chain's own rate is printed beside it every time so the number can only be read as the ratio it is.

CLUSTER.SYS — funding graph

Holders

Top holders examined-

Distinct operators behind them-

Largest single operator-

Funded with matching amounts-

Wallets under a day old-

Could it see enough to judge?

in the graph0 never seen0

A holder the graph has never seen is not evidence of innocence. Low coverage means the answer is weak, and it is shown rather than rounded away.

Clusters

-

OFFENDERS.DB

Creators with at least 15 launches, worst rate first. Sorted by rate and not by count on purpose: sorting by count just finds whoever is busiest, and the busiest address on any chain is usually a router everybody already knows about.

15 rather than 50 because the launch count on this chain is bimodal — measured over 167 live launches, the median creator has 3 and the 75th percentile has 719. Almost nobody sits in between, so any cut inside that gap picks out the same operators. Click any address to see everything else that wallet has done.

Creator	Launched	Pooled	Rate
loading…

INDEX.NFO

What this index holds

Launches indexed-

Distinct creators-

Reached a pool-

Got any liquidity-

Enough to trade against-

Anyone actually traded it-

Liquidity later pulled-

Every line above the bold one counts what the DEPLOYER put in. The bold one counts whether anybody wanted it: three quarters of funded pools were never traded, and a token nobody trades cannot be sold at any price. Two of every five funded pools held only dust, and more than a third of the rest had the money taken back out.

What it does NOT hold

Tokens with transfer history-

Tokens with holder lists-

Wallets in the funding graph-

Creator history covers every launch. The holder and funding layers cover only the tokens the desk examined closely, so most lookups get a creator read and an honest blank on the rest. Publishing that ratio is the point — a checker that hides its own coverage is asking to be trusted on the cases it never saw.

API.TXT

The same answer, as JSON

GET /api/token/<address>

GET /api/wallet/<address>

GET /api/gmgn

GET /api/creators

GET /api/index

Every field this window shows

There is no second, better answer behind a paywall. The console is drawn from the same payload the API returns, so what you can see you can automate.

Rate limiting, keys and the staked-access tier are not wired up yet — this build is open on localhost.

MINESWEEP.EXE

010 000

Left click opens, right click flags, double click on a number opens its neighbours once the flags around it add up. The first click is never a mine.

SIGNALS.LOG — calls the desk actually took

Standing

loading…

These are published while the position may still be open, so you are seeing what this desk holds before it has sold. Read them knowing that. Sizes, fills and P&L are not published — only the call and the reasoning behind it.

Buys — click a row for the full reasoning

Time	Symbol	Score	How each model graded it	Contract
none yet

GMGN.LOG — third-party feeds (observation, not calls)

Snapshot

Last look-

Creator record — the one thing here we have measured

Scan first — launchpad tokens people are also searching for

loading…

Launchpad (trenches)

loading…

Trending — and how much of it is just the hot-search list

loading…

Smart money & KOL — kept apart on purpose

loading…

SCORES.LOG — every verdict, newest first

The record, not the highlights

bought 0 skipped 0 conflicted 0

A ranking shows what the best looked like. This shows what the desk has actually been doing, which includes long runs of skipping everything — the most honest thing this feed can show you, and the part a top-25 list hides.

History — click a row for the per-model read

Time	Score	Symbol	Verdict	Kind	How each model graded it	What stopped it	Contract
nothing scored yet

TOPRATED.PNL — the best four models could find

What this is

Four independent models score every launch that clears the on-chain checks. This is the ranking, highest first, and it includes the ones the desk did NOT buy — a high score that was still passed on is the more interesting row, and hiding it would make the list look better than the record behind it.

loading…

Highest rated — click a row for the per-model read

Score	Symbol	Verdict	How each model graded it	Contract
nothing scored yet

README.TXT

What this answers

Who made this token, and what happened to everything else they made.

Why not a honeypot scan

Every token on this launchpad is minted from the same template, so a bytecode scan passes all of them and separates nothing. The contract is not where the risk is here. The creator is.

What it is not

Not a price prediction and not advice. A creator with a good record can still rug the next one, and a first-time creator is not guilty of anything — which is why a thin record is reported as thin rather than dressed up as a score.