ERC-8004 conformance census: four chains

354,858 agents. Base, BNB Chain, Ethereum mainnet, Celo, each pinned to a block.

This report supersedes 2026-07-29-base-cfbfcc01.md, which covered Base alone. "Supersedes" means later and wider, not republished: nothing in this project has been public. That report's numbers remain correct for the chain and run they describe, and §1 below is about what happens to their meaning once three more chains exist.

See METHODOLOGY.md for the rules behind every rung and CHANGELOG-METHODOLOGY.md for every correction this project has made to itself, including the ones in this report.

What this census found, in two facts:

Base attests at 49.2%. Across all four chains the rate is 12.2% — and Base, at 16.9% of the population, supplies 68.3% of every attested agent in existence.

Of those 354,858 agents, 358 have ever been paid and 34 have ever received a settlement through x402, the ecosystem's own payment protocol. One in 991, and one in 10,437. Checked against an independent index that has no concept of ERC-8004: of its 138 largest x402 sellers, none is a registered agent's declared wallet.

---

1. The correction this report exists to make

The Base report's headline was 49.2% attested. That number is correct. As a statement about ERC-8004 it was wrong, and only more chains could show it:

The spread is 44×. Any single chain, presented as "ERC-8004", is a claim about that chain's most active platform.

And the aggregate is barely better than the parts. 12.2% is the population-weighted rate, but it is not a fact about a healthy tail: 68.3% of all attested agents are on Base and 17.9% are on Celo, where — see §6 — three addresses wrote 99.8% of the attestations for the platform that owns 88% of the chain. Both numbers belong in the same sentence, every time.

The gas hypothesis is dead

The obvious explanation for the spread is transaction cost: attestation is a write, writes cost gas, so cheap chains should attest more. The data refuses it:

Base and BSC are both cheap L1/L2s and differ by 27×. Mainnet is the most expensive chain and outperforms BSC. Fee cost does not explain attestation, and no future report should reach for it. What does explain it, on the two chains examined closely, is the presence or absence of a single platform running an attestation service.

2. What was measured

Four runs, each pinning a block so that every agent's state is read from one simultaneous moment rather than assembled over hours:

The Identity Registry is at 0x8004a169… on all four, CREATE2-deployed to the same address; deploy blocks were measured by binary search on eth_getCode, verified on both sides of each boundary.

Two population footnotes, both from the runs' own manifests:

• BSC discovered 244,285 agents and swept 244,208. 77 could not be read from the chain. They are excluded from the run rather than recorded as failures — an RPC failure is the census's problem, not the agent's. They hold no feedback (§4 reconciles this exactly), so they cost nothing but the count.
• Base discovered 60,098 and swept 60,097. The one missing agent turns out to matter a great deal. See §4.

Rung 6 (live) is not implemented, so it is absent from every result rather than reported as skipped or fail. "We did not ask" and "we asked and got no answer" are different claims and the schema keeps them different. §8 says when it ships and why it has not.

3. Rung 2 mostly measures who inlines their document

The single most useful thing in the four-chain data is that a rung's meaning changed once there was something to compare against.

90.9% of Celo's agents put their whole registration document on-chain as a base64 data: URI. On Base it is 25.8%, mainnet 23.6%, BSC 59.6%. And an inline document passes rung 2 at 100%, on every chain, because there is no network involved:

This is correct, and the spec asks for it — line 52 permits any URI scheme and lines 192–195 recommend a base64 data: URI for fully on-chain metadata. Rung 2 asks whether the registration document can be retrieved; for an inline document the answer is yes by construction.

So: rung 2's cross-chain spread is largely a measure of how many registrants inline, not of how many run a working server. Celo's 97.4% and BSC's 74.0% are inlining rates wearing a reachability label. Where a document is fetched over the network, the four chains converge tightly — 25.7% to 36.3% — and that is the reachability finding: roughly two-thirds of agents that point at a server do not get a document back from it.

This is stated as a limitation of the measure, not a defect in the agents.

4. Reputation is machine-written, and Base's feedback is two-thirds unseen

Rung 7 counts feedback entries. Until now the census had never read one: the Reputation Registry returns a value with every entry and chain::Reputation::feedback kept only the count.

The values

Reading NewFeedback directly (analysis/feedback-values.md):

The largest single tag in the entire dataset — miner-vouch, 284,066 entries — has exactly one distinct value. Two tags recur across all four chains with near-identical values: trust = 85 (78.8–94.4% of its entries) and liveness = 100 (95.8–99.3%). BSC's entire feedback population is 104 addresses. And from rung 7's own evidence, 89.1% of BSC's attested agents have feedback from exactly one author (mainnet 88.8%, celo 74.3%, base 53.5%).

What this does and does not mean. An automated liveness prober writing 100 every time a check succeeds is behaving correctly and honestly — identical values are what a working automated check looks like. The finding is that the reputation layer is largely machine-written, not that it is dishonest. Nothing here re-derives any value or verifies that a trust of 85 reflects trust.

The one agent Base could not read holds 66.4% of its feedback

The event scan reconciles exactly with the census's own stored counts on BSC, Celo and mainnet — entry for entry. On Base it resolves to the digit:

143,713  what rung 7 recorded
+ 284,072  agent 25975
+      82  revoked feedback (getSummary omits it by design)
= 427,867  every NewFeedback event on Base

Agent 25975 holds 284,072 feedback entries from 62 clients — 66.4% of every feedback entry ever written on Base — and has no row in the run. It is the single agent the manifest already records as unreadable. The sweeper was right: a failed chain read is the census's problem, not the agent's, so the agent is excluded rather than failed, and the count was recorded durably. The failure was transient — ownerOf(25975) answers at the pinned block and now. What had never been done was joining that 1 to anything.

No published number becomes wrong. "Feedback on Base" as an unqualified phrase becomes unsupportable. Every Base feedback figure in this report and its predecessor covers 33.6% of the chain's feedback events, and says so wherever it appears.

5. The Validation Registry — 23 agents out of 354,858

ERC-8004 has three registries. The census reads two. The third had never been looked at: chains.validation_registry is NULL for every chain, which the code reads as "absent", and the validations table has zero rows.

Searching for the spec's own event topics with no address filter — because filtering on NULL returns zero and would have confirmed the assumption — finds that it is deployed, and used:

23 of 354,858 agents — 0.0065%. The mechanism has never been used at all on BSC or Ethereum mainnet: 285,014 agents, 80% of the population, zero events.

No canonical Validation Registry was ever deployed by the ERC-8004 team — the curated upstream says it is still under design. The ten that exist are third-party. Nine of the ten answer getIdentityRegistry() with the canonical Identity Registry this census sweeps, so these validations concern censused agents; the tenth reverts on that getter, which the spec requires at line 347, and its 10 events are reported separately rather than pooled.

A mechanism that has not been finalised has not been rejected — 105 requests is small, but "unused" and "unwanted" are different claims and only the first is measured. Full detail: analysis/validation-registry.md.

6. Celo is one platform, and it moves every chain-level number

Celo is the outlier on every measure: highest attestation (79.5%), highest resolvability (97.4%). It is one deployer's batch.

• Two EOAs hold 87.9% of the chain (6,934 and 1,630 agents), in adjacent contiguous mint ranges — ids 169–1,875, then 1,877–9,045 — with a largest single run of 1,503 consecutive agents. No other owner has a run longer than 72.
• Their documents self-identify as CeloNova, and that set is exactly those two owners.
• One of them emits 6,934 documents with a single top-level key shape (100%). Across rung 4, CeloNova produces one SHOULD-gap signature per 2,830 agents; every other owner, one per 43.
• Three EOAs wrote the feedback for 6,825 of CeloNova's 6,836 attested agents — 99.8%.

Limits, stated plainly. None of the three clients is either owner, and the Reputation Registry already forbids owner feedback. This census never reads ERC-721 approvals, so it cannot detect feedback from an approved operator — which the spec also bans (line 217) — and nothing here alleges it. Everything measured is equally consistent with a legitimate platform minting agents for its users, hosting their metadata on-chain, and running its own attestation service.

The finding is not about CeloNova's conduct. It is that Celo's 79.5% attestation rate is not an ecosystem-level fact — it is one platform's product decisions measured at chain scale.

Full detail: analysis/celo.md.

7. Concentration is the census's recurring result

Five independent measurements, five times the same shape:

Infrastructure — measured without probing anything

The rung-6 groundwork data, across all four chains, for agents that reached rung 4. 141,299 declared service entries:

11.0% of declared "endpoints" are not network endpoints at all — they are chain addresses, email addresses, empty strings, or absent.

The 125,705 HTTP(S) endpoints resolve to 3,399 distinct hosts:

Four hosts carry 59.2% of every declared endpoint in the census; eight carry 76.9%. Folding CeloNova's two hosts together makes it the largest single operator at 27.0%, and CeloNova plus evoevo.ai account for 47.9% of all declared endpoints across four chains.

An "open-ended agent economy" whose declared service surface is half two hostnames is a finding about the ecosystem, not about those two operators.

8. What this census does not measure

Stated so that absence is never read as evidence.

• Rung 6 (live) — whether any declared endpoint answers. Not implemented, therefore absent from every result, never a failure. It ships as the first follow-up report. It has not shipped because the method needs settling (402 is payment-gated, not dead; a 404 on GET may front a POST-only service; liveness ≠ functionality) and because the probe's User-Agent must first carry a domain that resolves and a mailbox that answers — currently it promises neither.
• Every other chain. Solana, Arbitrum, Polygon, Optimism and 25 more networks carry the same CREATE2 registries. Robinhood Chain (id 4663) has both registries deployed and zero agents minted. This bounds the cross-check in §9 too: 21 of x402scan's top sellers are Solana addresses, which our census cannot evaluate in either direction.
• Batch-settled x402. Only per-transfer EIP-3009 authorisations are visible. Any settlement aggregated off-chain is invisible.
• Funding graphs beyond two hops. "External" means the sender is not the agent's owner and not any owner in the run. An operator routing funds through one fresh intermediary is counted as external, so "external" is an upper bound on genuine earnings.
• Off-chain payments, and all non-stablecoin value. Native ETH, DAI, EURC and every other token are out of scope. Every payment figure is a lower bound.
• Feedback authenticity. Whether a value means what its tag says, and whether a client is independent of the agent. ERC-721 approvals are never read, so approved-operator self-feedback is undetectable.
• Which of seven parties "the agent" is. Owner, minter, approved operator, agentWallet, declared wallet, payment-contract controller, agent-controlled key and service operator can be eight different parties; the census reads three of them. See the role glossary.
• Intent. Ordering is not intent. That a wallet earned before its agent was registered says the registry counted activity rather than creating it; it does not say why.

9. Payments

Base's payment findings were retracted three times before publication — 313 agents paid became 190, $8,845,244 became $1,090,098, and "one operator earned 97.9%" was false. All three were one mistake in three costumes: an address treated as an identity. Full record: analysis/payments-corrections-ledger.md.

The funnels below apply those corrections from the start rather than retrofitting them, and their predictions were committed before any measurement was run (commit 8996609) so the result is a test rather than a rationalisation.

"Paid" means received an external stablecoin transfer after the agent was minted, through the agent's own declared wallet. Two stablecoins per chain, nothing else — every figure is a lower bound. Symbols and decimals were read from each contract: BSC's USDC and USDT are 18 decimals, not 6, and Celo's long-known cUSD address now reports USDm at 18.

Payment is three orders of magnitude rarer than registration. 358 agents of 354,858 — 1 in 991 — have ever been paid. 34 have ever received an x402 settlement: 1 in 10,437, and 32 of the 34 are on Base.

The thin-BSC prediction holds. BSC has 4.1× Base's agents and fewer paid ones, at a rate six times lower — on the chain holding 69% of all ERC-8004 agents.

Attestation does not predict payment. Celo attests at 79.5% and pays at 0.185% — a ratio of 430 to 1. Its attestation rate was three addresses writing feedback for one platform's batch (§6); it was never a measure of commerce, and the payment data shows what was underneath it. No report should let one of these measures stand in for the other.

x402 is a busy protocol that agents are barely part of. Base's stablecoins carried 6,875,861 AuthorizationUsed transactions in the blocks scanned; 8,904 reached an agent-declared address, averaging $1.07 — metering, not revenue. Mainnet's carried 26,260 and not one reached an agent. BSC has not a single such event on either token. Agents are absent from x402 on two chains for opposite reasons: on BSC the protocol is unused; on mainnet it is busy and agents are not in it.

Most of the value is not revenue. Contract-sourced flow dominates everywhere (50.8%–93.9%), and on Base the largest recipient's inflows are provably Morpho vault yield — the operator's own capital returning from DeFi.

Checked against a second source

The 34 is this census's most quotable payment number, so it was checked against an index built by someone else, by a different method, that has no concept of ERC-8004: x402scan (Merit Systems), which keys settlements by the seller's receiving address. Full working: analysis/x402scan-crosscheck.md.

Our 34 agents resolve to 22 receiving addresses. Of the 20 on Base, 18 corroborate — 9 of them to the exact settlement count. The 2 Celo addresses are outside x402scan's coverage: it indexes base, solana, polygon and optimism, not Celo, BNB Chain or Ethereum mainnet.

Two addresses diverge, and both resolve against their index rather than ours — but the hypothesis that would have gone the other way was tested first. Our x402 test is transaction-level, so a batching contract could have inflated us. It did not: for the divergent address all 1,679 flagged transfers sit in 1,679 distinct transactions, and Transfer.from equals the AuthorizationUsed authorizer in 8 of 8 sampled. The real cause is that x402scan's index is facilitator-scoped — 29 facilitators, 151 relayer addresses — while ours reads the chain directly; the relayer behind 93% of that address's settlements appears in none of them. In the opposite case x402scan reports 1,713 settlements for an address that has only ever received 1,667 USDC transfers in total, so its unit of count cannot be an incoming transfer.

No correction to our figures is forced. The 34 stands.

And the reverse direction, which is the stronger result

The same index answers a question our data cannot ask of itself: are x402's busiest sellers registered agents?

Taking x402scan's top 100 sellers by volume and top 100 by settlement count — 138 distinct EVM addresses — and testing each against our census with a deliberately generous rule (declared agentWallet, or the NFT owner, which is the spec's default value for agentWallet, on any of the four chains):

And all three sell on Base while owning an agent on a different chain — address reuse by an operator, not an agent being paid.

Agent-linked value is 0.017% of the top 100's volume, and our largest agent seller ($8,436) would not enter a list whose smallest member is $30,648.

Two independent datasets, joined on address, agree: the payment activity on x402 is not happening at registered ERC-8004 identities.

What this does not establish: a matching address confirms that a settlement occurred and who received it — never who initiated it or why. And a non-matching seller is not "not an agent"; it is not a registered ERC-8004 identity on the four chains we sweep. That is the finding: the registry is not where the payment activity is. x402scan publishes no disclaimer of its own — its repository contains no methodology page or statement of limits — so this caveat is ours and is not attributed to them.

The pre-mint correction is the largest single one. On Base, 82.5% of all value arriving at agent-declared addresses arrived before the agent existed — only 11.9% of transfers, but the early ones are far larger. Counting them is exactly how the retracted $8.8M was produced.

A fourth correction, caught in this round

Mainnet agent 28283 declares 0x0000…0000 as its agentWallet. The declared convention is unverified by construction, so the scan collected every USDC and USDT burn on Ethereum as income to that agent — 313,255 of mainnet's 314,735 transfers, 99.5%. Burn addresses are now excluded; Base and BSC have no such row, so no other figure moves.

That is PAY-1/2/3 in a fourth costume, and a fair measure of what the declared basis is worth: a services[] entry named agentWallet is a string in a document anyone can write, and one of them names an address that cannot hold anything.

What does not survive contact with the data

• "The agent economy." At 1 in 991 paid and 1 in 10,437 settling through the ecosystem's own payment protocol, the measured economy is a few hundred agents across four chains.
• Value totals as revenue. $3,009,509 is external post-mint inflow to agent-declared addresses. It is not earnings.

Limits

• "External" is an upper bound on earnings — the sender is not the agent's owner and not any owner in the run. Two hops, not a funding graph.
• Direction, not purpose. Airdrops, refunds and mistakes look identical.
• Zero means "nothing visible in scope", never "never earned anything".
• Agent counts are attributed through the declared map only, because the verified-agentWallet set could not be rebuilt from events (analysis/payments-per-chain.md §1). This understates every chain equally.

10. Reproduction

Every result above traces to one of the four run ids in §2, each of which stamps schema_version, checker_version, checker_commit, spec_commit and a literal rerun_command onto every row it wrote. The pinned spec was re-verified byte-identical to canonical ERC-8004 on 2026-07-30 (spec/SOURCE.md).

Event-derived figures — validation, feedback values, endpoint kinds — are reproducible with cast alone; each analysis document carries its own commands. Nothing in this report requires access to our database to check.

The x402 cross-check is reproducible without us or our method: x402scan's public tRPC procedures are unauthenticated and free, and analysis/x402scan-crosscheck.md carries the exact curl calls, the upstream commit (bf7a0cd), and the query date. Their documented REST routes are x402-paywalled at $0.01 each; no payment was made for anything in this report.