ASH REGISTRY $ASH

THE BURN LEDGER & SURVIVOR SCANNER — BOOST ERA, JUL 21 2026
opening the ledger…

Since 14:23 UTC every pump.fun graduation gets a five minute buyback, and every token it buys is burned. pump.fun shows you a bar while it happens. It never shows you the number. This registry reads that number straight off the chain — one billion minus what is left — then watches which coins were still standing once the buying stopped.

Burned by BOOST, counted MEASURED
TOKENS
Read off the chain across 0 BOOST windows. Every pump.fun coin mints exactly 1,000,000,000 and never mints again, so what is missing has been burned. That count is exact, and it only ever climbs. Those same tokens are worth / at this moment, and that figure falls as often as it rises: it is a valuation at today's market, not what BOOST paid. Most of these coins are worth a fraction of what they were when the buyback ran.
Network-wide ESTIMATE
SOL
graduations/hr observed on the live feed × 17.6 SOL per window, since Jul 21 14:23 UTC. Calibrating against the feed… This figure is inferred, not counted — the ledger on the left is the counted one.
0
In the ledger
0
Windows open
Survival, own windows
0
Filtered out

01Inspect a coin

Any Solana mint address. Answers in one read: did BOOST burn for it, and how much.
ASH — CONTRACT INSPECTION TERMINAL
Try: loading live examples from the ledger…

02Live windows

5:00 buyback window → 15:00 observation → verdict stamped. Scroll sideways.
Quiet. The next graduation on the feed opens a window here and the clock starts by itself.

03Survivor scanner

Who was still standing when the buying stopped.
# Token Graduated Burned Performance Holders Verdict ASH score
No verdicts on file yet. The first ones land ~20 minutes after a graduation.

ASH score /100: price hold after the window (0–35) + liquidity in dollars (0–30) + holder base (0–20) + burn weight against market cap (0–15), minus up to 20 for dev and concentration risk. Every input is an absolute figure, never a ratio of market cap — a coin whose cap collapses must not score better for it. Coins below the live-coin bar are not scored at all. COOKING ≥ +20% · HELD −10%…+20% · FADED < −10% · THIN not enough data. Performance marked OBS was measured by this registry from its own window-close reading; 1H is the venue's own hourly change, used for coins that graduated before we were watching. Those recovered coins reached the ledger because they were trading — a table of them will always look healthier than the truth, which is why the survival figure above counts only windows watched from the migration itself. A score is a measurement, not a call.

04Receipts

Printed at window close. Stamped at verdict. Newest first — scroll sideways.
The printer is idle. It never stays idle for long.

05How the burn is measured

No estimates in this section.

On July 21 at 14:23 UTC pump.fun stopped letting dead liquidity die. Every coin that graduates now gets a five minute buyback, and every token bought is burned. The migration transaction says so in plain text: an InitBoost instruction, then a Burn.

Burned tokens leave a permanent mark. Every pump.fun coin mints exactly 1,000,000,000 tokens and never mints again. So the burn is not something this registry has to guess at — it is one billion minus the current supply, readable by anyone, at any time, forever.

Here is the control, read live, same query for both halves. Coins that graduated before 14:23 UTC sit within a rounding error of their billion. Coins that graduated after it do not.

Graduated (UTC)AgoSupply nowMissing from the billion
reading supplies…

Live reading. The top rows are BOOST-era graduations; the bottom rows graduated before BOOST went live.

There is one more check, and it is the one that matters. pump.fun states the buyback is ~17.6 SOL per SOL-paired window. This registry never uses that figure to produce a burn — it reads supply. So the two are independent, and what the ledger measures can be held against what was announced.

One caveat the registry will not paper over: a supply deficit says tokens were burned, not who burned them. Anyone can burn tokens they own, and pre-BOOST coins do carry small deficits from ordinary holder burns — tens of thousands of tokens, against the hundreds of thousands to tens of millions BOOST takes out. So a window has to clear 25,000 tokens before it counts here, and inside the BOOST era nothing else burns at that scale.

The rest follows from that. A receipt is printed when a window closes, with the supply reading, the market cap, the liquidity and the holder count at that moment. Fifteen minutes later the same coin is read again and stamped. Nothing on this page is invented; where a number is inferred rather than counted, it is labelled EST. and kept away from the ledger.

The keeper predicts nothing and advises no one. It counts. When $ASH is live, its own furnace gets a page here — creator fees in, burns out, every transaction attached.

PumpPortal
Live migration feed. Opens a window the second a coin graduates.
Jupiter
Supply, holders, liquidity, dev audit. The supply reading is what the burn is measured from.
GeckoTerminal
Second pair of eyes on new PumpSwap pools, so a dropped socket does not mean a missed window.
DexScreener
Pair-level cross-check on anything you inspect by hand.