⌘K

Live evidencecounted 18m ago

0.75days of PURR volume to unwind the crowd98%of the XMR book on one side, across 33 vaults$1.60Bunder one key with no guardian, 100 of the 481 vaults whose control we have read$347.1Mwhose control we have never read, across 3,321 vaults$1.95Bsitting in vaults and not lent out$11.61Bunder 3,802 vaults we read$4.68Bin 331 vaults where we saw automated execution, nothing declared2,199vaults declared or on a venue built to run software, holding $1.11M333software-run vaults scored, 13% of them456operators identified behind them26networks where we read capital, from Ethereum to Cardano32,011,310deposits and withdrawals read from chain147changes recorded in the last day916,673ERC-8004 agent registrations read+2.54/ 30 days679,104distinct owners behind those registrations203,350registrations carrying a readable card, 22%3,390,306paid requests on the Olas rail (deliveries are counted apart)−0.5/ 30 days88suppliers delivering that work412,718agent-to-agent jobs read on Base+0.53/ 30 days$1.0Msettled between agents, all time1,176suppliers listed on the ACP rail187agents live on Cardano78Cardano agents deregistered, 29% of all ever minted29,745escrows between Cardano agents since Jun 2025, 157 still open78%of every agent payment we read comes from 2,927 addresses that never stop339,663addresses paid once and never came back, of 572,437 that ever paid1,080of those payers are an agent we can name, 0.189% of them683agent hosts answered when we asked, of 2,579 probed17,115agents sit behind a host that answers nothing at all458msmedian answer from the 683 hosts that answer42%of all agent ratings were written by ten addresses, of 735,701 ratings read75,042rated agents have exactly one rater, of 93,000 rated at all116,984service endpoints declared across 50,269 agent cards we read0.75days of PURR volume to unwind the crowd98%of the XMR book on one side, across 33 vaults$1.60Bunder one key with no guardian, 100 of the 481 vaults whose control we have read$347.1Mwhose control we have never read, across 3,321 vaults$1.95Bsitting in vaults and not lent out$11.61Bunder 3,802 vaults we read$4.68Bin 331 vaults where we saw automated execution, nothing declared2,199vaults declared or on a venue built to run software, holding $1.11M333software-run vaults scored, 13% of them456operators identified behind them26networks where we read capital, from Ethereum to Cardano32,011,310deposits and withdrawals read from chain147changes recorded in the last day916,673ERC-8004 agent registrations read+2.54/ 30 days679,104distinct owners behind those registrations203,350registrations carrying a readable card, 22%3,390,306paid requests on the Olas rail (deliveries are counted apart)−0.5/ 30 days88suppliers delivering that work412,718agent-to-agent jobs read on Base+0.53/ 30 days$1.0Msettled between agents, all time1,176suppliers listed on the ACP rail187agents live on Cardano78Cardano agents deregistered, 29% of all ever minted29,745escrows between Cardano agents since Jun 2025, 157 still open78%of every agent payment we read comes from 2,927 addresses that never stop339,663addresses paid once and never came back, of 572,437 that ever paid1,080of those payers are an agent we can name, 0.189% of them683agent hosts answered when we asked, of 2,579 probed17,115agents sit behind a host that answers nothing at all458msmedian answer from the 683 hosts that answer42%of all agent ratings were written by ten addresses, of 735,701 ratings read75,042rated agents have exactly one rater, of 93,000 rated at all116,984service endpoints declared across 50,269 agent cards we read

Agent work

Read 17 minutes ago

What agents buy from other agents, settled on chain.Note 23

Finding

provider addresses delivered work

35.0%

of all work from one of them

72.0%

from the top three

3,390,306

request-side payment events

943,911

delivery events

Olas Mech · gnosis 98%, base 2%, optimism 0.0%, polygon 0.0% · 2025-07-07 to 2026-10-20 · 471 days

A market a competition authority would call highly concentrated.

How this is counted

The last two are counted separately and not divided. Nothing readable links a request to the delivery it became, so a ratio between them would look like a completion rate and would not be one, and it would disagree with the venue’s own lifetime figures, which a reader would be right to trust instead.

Who supplies it

concentration 2,199 HHI · highly concentrated

Provider	Network	Tasks delivered	Share of all work	Fees earned
0x601024e2…151f	Gnosis	330,602	35.0%	8190.10 xDAI
0xc05e7412…632c	Gnosis	247,241	26.2%	9105.31 xDAI

The same question, read over four windows

why the figure above is the bottom row

Read over	Suppliers	Deliveries	Top one	Top three	HHI	Would read as
30 days	37	55,639	42.9%	70.0%	2,257	highly concentrated
90 days	45	143,145	44.4%	71.2%	2,405	highly concentrated
180 days	49	387,208	45.9%	77.2%	2,761	highly concentrated
all of it	88	943,911	35.0%	72.0%	2,199	highly concentrated

Is the tail real?

The windows above cannot say. An HHI of two thousand is produced both by three large suppliers with a hundred small ones behind them and by three with nothing behind them, and a buyer is choosing between those two markets.

Every concentration figure this site published was once read over the shortest window here. Widening it multiplied the supplier count 2.4× on this rail and moved the top supplier’s share from 42.9% to 35.0%. We show the windows rather than the best of them, because a row marked as disagreeing with the full history is a row that would have had us tell you something else.