SCIENTER

Measured record

Proof Board

This is the measured record of one published rule, applied to every pump.fun launch our recorder saw. Each decision was written into a hash-chained log within seconds of the launch, before anyone knew how the launch would end. The outcome was read 24 to 30 hours later. It is a record of what happened, not advice about any token.

Integrity check failed for 2026-11-02, 2026-11-03, 2026-11-04, 2026-11-05, 2026-11-06, 2026-11-07, 2026-11-08: a day's closing hash was not written in time; a day's board could not be made from its files. Those days are out of every total until it is explained here.
Updated 2026-11-19 07:05 UTC. Made with proofkit-1-d2b3ebe36d25f017. Roots saved outside Scienter: the Internet Archive lists 10 copies of this page, the latest taken 2026-11-19 01:17 UTC. Each holds whatever this page showed when it was taken. See the list at the Internet Archive

1

The record so far

The rule (deployer-all-priors-died/1): the wallet that launched the token had at least one earlier launch with a known outcome, and none of those earlier launches was still trading.

Nothing is counted yet. Counting starts with launches from 2026-11-18; the first of those days is complete after 2026-11-26. Until then every day is shown below and none is added up.

45,411 flags committed so far, out of 205,490 launches, since 2026-11-02.

2

Every day

The pre-registration was planned to freeze on 2026-11-10; the chain first carried its hash at 2026-11-09 18:37 UTC; counting starts with launches from 2026-11-18.

The last four columns are the percent still trading at the check, for that day alone.

Every launch day, newest first
Launch dayStatusLaunchesFlagged launchesChecked on timeChecks requestedFlagged by the ruleNot flagged: the wallet had an earlier launch still tradingNo record: the wallet's first known launchAll launches
2026-11-18open, outcomes still coming in11,1954,4070417
2026-11-17open, outcomes still coming in12,3733,68012,3213670.44%1.57%1.35%1.09%
2026-11-16open, outcomes still coming in11,0163,70710,9752720.46%2.38%1.45%1.18%
2026-11-15open, outcomes still coming in8,1722,2938,1452810.39%3.43%1.34%1.20%
2026-11-14open, outcomes still coming in8,3832,7038,3613830.63%4.04%1.86%1.59%
2026-11-13open, outcomes still coming in11,5082,86211,4763870.70%5.34%1.42%1.46%
2026-11-12open, outcomes still coming in11,6302,55611,6007380.47%2.03%2.94%2.35%
2026-11-11open, outcomes still coming in12,2882,73012,2507320.37%2.87%3.76%2.97%
2026-11-10first week, not counted13,8474,05413,8066730.17%2.40%3.14%2.25%
2026-11-09before the rule was frozen, not counted12,2891,89312,2314200.90%5.04%3.09%2.81%
2026-11-08before the rule was frozen, not counted11,0981,4126,671not published0.68%3.23%1.89%1.80%
2026-11-07before the rule was frozen, not counted11,8401,7970not published
2026-11-06before the rule was frozen, not counted15,0772,2560not published
2026-11-05before the rule was frozen, not counted13,7922,6860not published
2026-11-04before the rule was frozen, not counted17,1022,7070see the file
2026-11-03before the rule was frozen, not counted15,5802,4030see the file
2026-11-02before the rule was frozen, not countedre-boarded under proofkit-1-d2b3ebe36d25f017; earlier board: board-20261102-353b360c9dc1f8a8.json8,3001,2650see the file

11,311 launches were recorded twice by the recorder; the second record is not counted anywhere.

3

Flagged, still trading at the check

Each of these was flagged at launch and was still trading at the check, with the liquidity DexScreener showed. A float of $1,000 held for a few minutes around a known check time is enough to appear here, so read the liquidity column.

Flagged launches that were still trading at the check
Launch dayTokenLaunching walletLiquidity at the checkChecked at
2026-11-106r7D6…hpump8AcZB…pdQmt$2,0352026-11-11 06:42 UTC
2026-11-10E8Kkw…FLcy2timew…2Lsyn$2,0232026-11-11 08:30 UTC
2026-11-10DD3ca…vM3HYDfv8M…eAUJG$2,5962026-11-11 11:00 UTC
2026-11-106fXnQ…4pump47zTd…mYGaY$3,0432026-11-11 17:48 UTC
2026-11-10BhJnv…xpump45e2N…YaXoT$1,8042026-11-11 22:48 UTC
2026-11-10Buf5o…kpumpArK1Z…wRYQe$15,4022026-11-11 23:20 UTC
2026-11-104CiZ8…MpumpFLoCn…meXPY$6,8602026-11-11 23:49 UTC
2026-11-09e9YfL…648WTtimew…2Lsyn$10,6692026-11-10 05:03 UTC
2026-11-099w2NE…jpump862TY…RGWo2$2,7472026-11-10 06:51 UTC
2026-11-095Sdgg…cswurpopsP…E8kxS$1,5192026-11-10 09:30 UTC
2026-11-09FoMrt…1S145popsP…E8kxS$1,5732026-11-10 09:55 UTC
2026-11-09FyKEr…B7uo3popsP…E8kxS$1,5232026-11-10 10:09 UTC
2026-11-09BD2TC…NDn1wpopsP…E8kxS$1,7102026-11-10 10:09 UTC
2026-11-09GEbmi…ALPvEpopsP…E8kxS$1,5232026-11-10 10:14 UTC
2026-11-0988QyG…p61kapopsP…E8kxS$1,4962026-11-10 10:14 UTC
2026-11-0976AF4…csBh8popsP…E8kxS$1,5352026-11-10 10:21 UTC
2026-11-095916J…npump8AcZB…pdQmt$3,1242026-11-10 10:38 UTC
2026-11-09FCxfR…LpumpB8oEj…8kcxm$2,1902026-11-10 13:16 UTC
2026-11-094aqpn…9pump241nH…jBny2$2,2952026-11-10 13:39 UTC
2026-11-09GjTxR…VpumpDYmsQ…9TPNF$2,7792026-11-10 13:59 UTC
2026-11-093iEBq…jpump4M47u…6yTgX$1,7452026-11-10 19:37 UTC
2026-11-09Dr1Yv…Lpump3F8m5…YhBQs$2,9002026-11-10 20:54 UTC
2026-11-09FBeSu…2pumpDYmsQ…9TPNF$2,9892026-11-10 21:05 UTC
2026-11-09Ff3aS…NFLKVBT9Ph…wzNwd$1,9742026-11-10 21:13 UTC
2026-11-084wQ3a…Y1rcq9wRuF…xTFRG$2,0712026-11-09 17:18 UTC
2026-11-086v1E8…9pump4rjcg…uubh2$1,6932026-11-09 17:19 UTC
2026-11-087ETtD…Xpump862TY…RGWo2$8,6882026-11-09 17:21 UTC
2026-11-087gK4Y…Ypump9VhXE…1cYz4$11,9232026-11-09 17:21 UTC
2026-11-08DXoFo…5pumpFcrfA…McfKF$2,0312026-11-09 19:20 UTC
2026-11-084tan3…7JxdQpopsP…E8kxS$1,5142026-11-09 19:48 UTC

4

What the numbers mean

Still trading means DexScreener listed a pair for the token with at least $1,000 of liquidity at the check. No longer trading means it did not.

A rate is the share still trading, then the counts it comes from, then a 95% interval in brackets. Where a group has no launches checked on time the rate is left empty.

An interval marked counted by wallet treats all launches by one wallet as a single unit, so a few busy wallets cannot narrow it.

Launches is every launch the recorder saw that day. Checked on time means the outcome was read 24 to 30 hours after the launch.

A day is counted when the rule was already frozen, it is past the first week, it is complete, every group was checked on time for at least 90% of its launches, no hour of launches was under half checked, its lines were written promptly, and its overall rate did not move sharply against recent days. Totals add up counted days only.

Most launches stop trading whatever the record. Read the flagged row next to the other rows, never alone.

5

Verify it yourself

You need Python 3.10 or newer and about ten minutes. No account, no key.

  1. Download proofkit.py, the pre-registration, one day's log, and SHA256SUMS. Check the file's SHA-256 against SHA256SUMS.
  2. Run: python proofkit.py verify radar-YYYYMMDD.jsonl.gz --prev <the previous day's root>. It checks every line's hash, the link to the previous day, and that the last hash equals the root published for that day.
  3. Run: python proofkit-<kit>.py board --day YYYYMMDD with that day's file and the eight after it, using the verifier named in the board's kit_version. The output must match board-YYYYMMDD-<kit>.json byte for byte.
  4. Open any flag: /api/proof/event/<seq> shows the line, its hash recomputed, and its day's root.
  5. Save today's root somewhere we do not control. That copy is what makes a later rewrite detectable.

The same code made this page and the files you check it against. Matching bytes prove the page came from the log; they do not prove the rule is right. Read the code.

Files

Published roots

The last hash of each UTC day's log, published when the day ends, and the chain head every hour.

  • 2026-11-19 07:05 UTC · seq 409680

    0e4a73d1945b6e85d4d13e285cc2568abecdf5e76a17807fdfdf71d97b5ec3c2
  • 2026-11-18 · seq 403647

    a6d7da086a2f8ff27fe1084976dd6dfeca33992aa4c3ba98bb3b2f71532020bb
  • 2026-11-17 · seq 379932

    1b8f3b05accc7645bb7338fed2dedb06964c2e529db879964abdee85ef8f3b03
  • 2026-11-16 · seq 356480

    07a75974b99a12d21f9814ee16aee67de8d3975a48aadccc95a10c12e02f97f9
  • 2026-11-15 · seq 337165

    c0087ff08929361697dc69edd0e5a62d64ab5fd8a6d2009929de9056488e03bf
  • 2026-11-14 · seq 320464

    ca2dcddcb770978e1ff5b295eaba4ed02c0ab6abf62329d72c328274618e71fa
  • 2026-11-13 · seq 300452

    0fa4001a486e27188bef5962ca6b56270a2eca385b6cc72a8fd6dace1fe0b9ae
  • 2026-11-12 · seq 277104

    e7b79ba6a9f238597eaa602b1c2855ba737c3b84654f80732c076530f5f6976b
  • 2026-11-11 · seq 243089

    2cbcc2b21f85d0bc59447fb1928f521b6e70ba3fd8149a45d10ab572419a31ad
  • 2026-11-10 · seq 208579

    e933001106fc218d9e2699c00979e7da39f113618af8f1503dba6d9628374aaa
  • 2026-11-09 · seq 174334

    f3ed2d44e67cc01be7f0c5c04bb666466f08ed52737ec65f327d3c647be7847a
  • 2026-11-08 · seq 149225

    b31b6a83347605c1646dd17b9ec8f55051a53d47fd35b7208a30e7d0153e6987
  • 2026-11-07 · seq 119858

    fa7e9e5edb5ce082e5bd684b70b8c535d4a22567f52571e71da7c884fdc1ebd9
  • 2026-11-06 · seq 89737

    6c6bbd19c07b3b330a0f0fbcc7b5cacf0df2e0ca75f8160963497fca84a7468f
  • 2026-11-05 · seq 56340

    3a71cec79ff0280b78ee853d1d57654dacecf07532bc0b1cd934d692692a73a6

6

What this does not show

  • Until someone other than us saves a root, the record rests on our word: we could rewrite the log and publish new roots. Any saved root exposes that.
  • We do not store DexScreener's reply from the time of the check. You can re-check a token today, not replay the check.
  • Tokens DexScreener never listed count as no longer trading, in every row alike.
  • Liquidity can be added just for the check; it costs little. Every flagged launch that was still trading is listed above with its liquidity.
  • Timestamps, and the date the rules on this page were written down, are ours until a third party saves a root from that time.
  • The record each decision rested on is our index at the moment of the launch; it cannot be rebuilt from outside.
  • How many launches the recorder misses has not been measured yet.
  • Launches our recorder missed are not here.
  • From line 156184 the log names the outcome source correctly; earlier lines carry an older label for the same check.