The Trust Layer the Protocols Assume Is Already There: A Second Independent Paper Puts Our Live Data In Its Empirical Corner

A June 2026 protocol-governance survey (arXiv:2606.31498, Kang & Diponegoro, DoiT International) explicitly scopes reputation out of its six-dimension governance taxonomy on the grounds that it is "already addressed by ERC-8004." An independent empirical study of that same registry (arXiv:2606.26028, Imperial College London) finds that on deployed chains the ReputationRegistry "cannot function as a trust signal." The two papers are not strictly contradictory — one is a taxonomy scoping decision, the other a measurement of deployed state — but read side by side the honest framing is the assumption is reasonable on paper and false on chain. Our live continuous index — three months fresher than either paper — is the tie-breaker.

Published 15 August 2026 · Every headline count describing our own index is pulled live from /v1/public/stats at build time and refreshed on every visit. Third-party figures from arXiv:2606.31498 (Kang & Diponegoro, 30 June 2026) and arXiv:2606.26028 (Imperial College London, 24 June 2026) are frozen citations dated to each paper's publication and pinned by test so they cannot drift silently. Third of three independent-source posts published today alongside the x402 concentration reconciliation and the Imperial College reconciliation.

The AWS Bedrock AgentCore observation: the largest cloud vendor shipped the same gap

We want to lead with what arXiv:2606.31498 Section V.C describes, because it moves the argument from the specification stack into deployed production infrastructure. Kang and Diponegoro note that AWS Bedrock AgentCore — GA in 15 regions — ships a production agent registry with semantic search, A2A Agent Cards, Cedar-based authorization, lifecycle management, and CloudTrail audit; and that this registry "encodes no trust scoring, behavioral reputation, capability verification, or governance primitives." Their own summary of the implication is that "the governance gap is not merely a research finding; it affects deployed enterprise systems serving production agent fleets."

That is independent confirmation from the largest cloud vendor in the category that the trust layer is unbuilt. AgentCore is not a research artifact; it is the discovery-and- registry substrate underneath a wave of enterprise agent deployments. If AWS's own production agent registry ships without behavioral reputation or capability verification, the trust filter is genuinely a missing architectural layer, not a feature that everyone else has and we somehow overlooked.

The governance gap matrix — frozen third-party citation

Kang and Diponegoro publish a six-dimension governance taxonomy (membership, deliberation, voting, dissent preservation, human escalation, audit/replay) and score each of the five major agent interoperability protocols against it. Each protocol–dimension pair is classified Supported (2 pts), Partial (1 pt), or Absent (0 pts) with a coverage denominator of 12. The published gap matrix (their Table III, dated 30 June 2026) is:

Protocol	Coverage (out of 12)
MCP v1.1 (Anthropic; tool access)	1/12
A2A v1.0.1 (Google / Linux Foundation; delegation)	1/12
ACP (IBM Research; communication)	2/12
ANP (DID-based routing)	0/12
ERC-8004 (on-chain identity, reputation, validation)	2/12

Source: arXiv:2606.31498 Table III (Kang & Diponegoro, 30 June 2026). Voting, dissent preservation and human escalation are Absent across all five protocols. Frozen citation — do not extrapolate the classifications to a later spec revision without re-reading the paper against the current spec.

The two papers, side by side — a scoping decision and a measurement

The reconciliation is not "one paper is wrong." Both papers are correct within their scope; they are measuring different things. The honest framing is that the assumption Kang & Diponegoro make on paper about the trust layer is false on chain today — which is a statement about the deployed state, not about the taxonomy.

Source	What it does	What it says about reputation / ERC-8004
arXiv:2606.31498 Kang & Diponegoro, 2026-06-30	Specification-level gap analysis of five agent interoperability protocols against a six-dimension governance taxonomy (G1–G6). Scoping decision about what belongs in the taxonomy.	Reputation is "a prerequisite for governance ('who is credible?') but not itself a governance primitive—it is already addressed by ERC-8004." Reputation is therefore scoped out of the governance taxonomy and treated as a solved neighbouring problem.
arXiv:2606.26028 Xiong et al. (Imperial College London), 2026-06-24	Empirical study of the ERC-8004 identity + reputation registries as deployed on Ethereum, Base, and BSC. Measurement of deployed state.	Only 3% (Ethereum), 4% (BSC), and 15% (Base) of ERC-8004 registrations satisfy the paper's trust bar. In their own words, "most registrations are placeholders rather than active agents," and the `ReputationRegistry` "cannot function as a trust signal."

The papers are not contradictory. Kang & Diponegoro make a scoping call about what a governance taxonomy should include; Xiong et al. measure what the trust registry currently looks like on chain. Together they say: the trust layer everyone else's protocols assume is already there is a real architectural assumption — and it is not currently held up by what is deployed.

Our live numbers, three months fresher than either paper

The registry Kang & Diponegoro scope reputation out to is the same one Xiong et al. measured on 2026-05-13 and found wanting. We index it continuously. On this page load, the live counts are:

25,618 ERC-8004 agents registered on Base + Ethereum mainnet.
3,323 answer at their advertised endpoints (Base 7%, Ethereum mainnet 58%).
277,710 ReputationRegistry feedback events indexed, of which 44 agents have rows that tie back to a real, paid on-chain job (ERC-8183 / Virtuals ACP outcome or an allowlisted hook contract).
Liveness is one code path (smartcontractauditpro/liveness.py::is_live) continuously tracked since 2026-07-17.

The three-month-fresher continuous view lines up with the Imperial College snapshot: Base has the highest live-rate, Ethereum comes second, and BNB (which we no longer track by deliberate choice and which Imperial had at 4%) is the weakest of the three chains under both teams' predicates. Full side-by-side of both predicates is at our reconciliation of the Imperial College paper. Full supply-side treatment of the same finding from the x402 side is at the x402 concentration reconciliation. Report 07's live strong-filter cohort number is on /blog/x402-demand-concentration-2026-08.

Why this matters: the trust filter is a missing layer, not a feature

Every protocol in Kang & Diponegoro's matrix assumes some other layer solved trust. MCP is tool-centric — not its problem. A2A is delegation-centric — not its problem. ACP is communication-centric — not its problem. ANP is routing-centric — not its problem. ERC-8004 is trust-centric — scopes governance out. AWS Bedrock AgentCore ships production without trust scoring or reputation. Kang & Diponegoro's own Section V.A observes that A2A's extension mechanism has been publicly available for "6+ months ... with an active extension ecosystem [and] zero governance extensions have been proposed or implemented." That is the ecosystem-side signal: the gap is recognised in the specification stack and remains uncontended by the community that would extend it.

Which is exactly the shape our other two independent-source posts land on from the payments side. On the x402 supply side, the top ten services take almost all the settled volume (two-source reconciliation). On the ERC-8004 identity side, most registrations are empty (Imperial College reconciliation). And now, from the protocol-governance side, the community explicitly assumes some other layer solved trust. Three orthogonal readings of the same underlying fact: the trust filter that everyone assumes exists actually needs to be a separate architectural layer somebody builds.

What we're doing about it

Our answer to Kang & Diponegoro's assumption and Imperial College's measurement is the same: /commerce-backed-agents — the free, indexable hub of every ERC-8004 agent whose ReputationRegistry row ties back to a real, paid on-chain job. Canonical predicate lives at smartcontractauditpro/commerce_backed.py, with a divergence-regression test keeping the per-agent classifier and the SQL predicate in agreement. Underlying the cohort is a live continuous index — 25,618 agents on Base + Ethereum mainnet, of which 3,323 answer at their advertised endpoints. The commerce-backed cohort is what a real "trust filter" architectural layer looks like today — the small cohort of agents that a real buyer has paid, more than once, on-chain, in a window that a bulk-mint farm cannot cheaply reproduce. It is the single signal the registry does not carry natively and that no protocol in the Kang & Diponegoro matrix encodes.

On the API side, the same signal is a paid field on /v1/intel/agent/{agent_id}: the Sybil-adjusted reputation score, commerce-backed status and job-outcome count, hook-type classification, and the by-service revenue breakdown. Payable per call in USDC or ETH via x402, or unmetered on a Pro subscription. That is what it takes to have a trust filter that survives the gap Kang & Diponegoro assumed had been filled and Xiong et al. measured empty.

Attribution and reuse

Quotations from arXiv:2606.31498 (Richard Kang & Yudho Diponegoro, Governance Gaps in Agent Interoperability Protocols: What MCP, A2A, and ACP Cannot Express, 30 June 2026) are used under fair-use for critical commentary and educational purposes. The paper is distributed under the arXiv perpetual non-exclusive licence; short quotations and paraphrase only, full attribution and link to source above.

Quotations from arXiv:2606.26028 (Xihan Xiong, Zelin Li, Wei Wei, Qin Wang, William Knottenbelt, Zhipeng Wang — Imperial College London, Can Trustless Agents Be Trusted? An Empirical Study of the ERC-8004 Decentralized AI Agent Ecosystem, v1 2026-06-24, v2 2026-07-08) are used under CC BY 4.0.

Our own live numbers on this page are served from /v1/public/stats and refresh every 15 minutes. Redistribution of our figures is welcome under CC BY 4.0 with a link back to this page.