⌘K

Live evidencecounted 7m ago

0.51days of MEME volume to unwind the crowd99%of the XMR book on one side, across 28 vaults$1.28Bunder one key with no guardian, 52 of the 424 vaults whose control we have read$315.2Mwhose control we have never read, across 3,225 vaults$1.96Bsitting in vaults and not lent out$11.34Bunder 3,649 vaults we read$4.87Bin 323 vaults where we saw automated execution, nothing declared2,080vaults declared or on a venue built to run software, holding $1.01M325software-run vaults scored, 14% of them451operators identified behind them26networks where we read capital, from Ethereum to Cardano31,241,990deposits and withdrawals read from chain19changes recorded in the last day588,846ERC-8004 agent registrations read−0.6/ 30 days360,982distinct owners behind those registrations199,780registrations carrying a readable card, 34%3,330,722paid requests on the Olas rail (deliveries are counted apart)−0.67/ 30 days87suppliers delivering that work411,027agent-to-agent jobs read on Base+0.05/ 30 days$1.0Msettled between agents, all time1,175suppliers listed on the ACP rail189agents live on Cardano76Cardano agents deregistered, 29% of all ever minted29,009escrows between Cardano agents since Jun 2025, 144 still open78%of every agent payment we read comes from 2,571 addresses that never stop307,376addresses paid once and never came back, of 494,774 that ever paid934of those payers are an agent we can name, 0.189% of them635agent hosts answered when we asked, of 2,532 probed17,125agents sit behind a host that answers nothing at all404msmedian answer from the 635 hosts that answer44%of all agent ratings were written by ten addresses, of 671,178 ratings read29,502rated agents have exactly one rater, of 47,110 rated at all86,979service endpoints declared across 33,840 agent cards we read0.51days of MEME volume to unwind the crowd99%of the XMR book on one side, across 28 vaults$1.28Bunder one key with no guardian, 52 of the 424 vaults whose control we have read$315.2Mwhose control we have never read, across 3,225 vaults$1.96Bsitting in vaults and not lent out$11.34Bunder 3,649 vaults we read$4.87Bin 323 vaults where we saw automated execution, nothing declared2,080vaults declared or on a venue built to run software, holding $1.01M325software-run vaults scored, 14% of them451operators identified behind them26networks where we read capital, from Ethereum to Cardano31,241,990deposits and withdrawals read from chain19changes recorded in the last day588,846ERC-8004 agent registrations read−0.6/ 30 days360,982distinct owners behind those registrations199,780registrations carrying a readable card, 34%3,330,722paid requests on the Olas rail (deliveries are counted apart)−0.67/ 30 days87suppliers delivering that work411,027agent-to-agent jobs read on Base+0.05/ 30 days$1.0Msettled between agents, all time1,175suppliers listed on the ACP rail189agents live on Cardano76Cardano agents deregistered, 29% of all ever minted29,009escrows between Cardano agents since Jun 2025, 144 still open78%of every agent payment we read comes from 2,571 addresses that never stop307,376addresses paid once and never came back, of 494,774 that ever paid934of those payers are an agent we can name, 0.189% of them635agent hosts answered when we asked, of 2,532 probed17,125agents sit behind a host that answers nothing at all404msmedian answer from the 635 hosts that answer44%of all agent ratings were written by ten addresses, of 671,178 ratings read29,502rated agents have exactly one rater, of 47,110 rated at all86,979service endpoints declared across 33,840 agent cards we read

What changed

Material changes across the corpus. Raw activity is collapsed into economically meaningful events.Note 20

51 days of this feed, from when we started recording it, from 19 August 2026. That is short, and it is short because we started recording it then rather than because nothing happened before. This is a reading we took and stored rather than an event a chain kept, so the series itself begins when we began; the underlying activity may well be reconstructable from an archive node or the venue’s own history, and we have not done it.

Pulse

Events

in the last 24h

$25.43M

Capital behind them

summed across events that carry a figure

Subjects touched

vaults and instruments

Above $10M

events large enough to matter alone

All 19 Capital 17 Agent work 2 Identity 0|Flows 9 Lifecycle 4 Risk 3 Jobs & settlement 2 Crowding 1|24h 7d All

19 of 19 shown · recording since 2026-08-19

24 kinds of event

asset repriced9
sources disagreed38
curator changed143
guardian removed47
guardian set46
control.owner7
became crowded328
we corrected a denomination1
exitable again63
stopped being exitable63
money in1371
money out1326
ratings written153

What became of it

Every adverse event on the tape, against whether the opposite event followed it on the same vault. Only the kinds that HAVE an opposite are here: a crowd does not announce that it thinned, so crowding is on the feed above and cannot appear in this table.

Event	Happened	Came back	Median to recover	Still open
Risk rose	2,445	1,795	1.5d	650
Stopped being exitable	63	61	8h	2
Confirmed dead	18	0	—	18

Measured over the whole register rather than the window above, because a recovery that arrives after the window would otherwise be counted as a failure to recover. A dash under the median means nothing in that row has come back yet.

The ones that have not come back

200 with nothing since, largest first

Vault	What happened	Open for	Holds
Adpend USDC	Risk rose	24d

What it will and will not tell you

A change on a vault holding under $50,000 is not announced.: The register still carries every row. This decides what is worth interrupting somebody for, and a feed that reports a test deployment beside a hundred-million-dollar lock has told the reader the two are comparable.
Recovery is reported as loudly as failure.: A vault becoming withdrawable again is the half an alert feed tends to drop, because it is not alarming. It is also the half that tells you whether a lock was a squeeze or a failure.
Each change is said once.: A vault that stays locked is one entry, not one every four hours. The failure of a feed is never a missed event, it is repeating itself until nobody reads it, at which point the one that mattered goes past unread too.
Nothing before the feed started is here.: It was not missed. It was not yet being watched, and inventing it afterwards is not available to us.