Sheet RDKG / FS-07 Lane Professional — Verified Foresight Drawn 2026-09-02 Rev v0 · draft Gate Founder review

The Grounding
Protocol

How humans and machines talk, how machines talk to each other, and the one connector that is missing from both.

← Watch the film (1:48)

“Do not say that for which you lack adequate evidence.”

Grice, Logic and Conversation, 1975 — the oldest rule in communication, and the only one that has never been given a wire format.

This sheet surveys both conversations — person ⇄ model and agent ⇄ agent — across the canon and the 2026 frontier. They turn out to be failing at the same thing, in the same place, for the same reason. The last three plates propose the connector, name it, and cost it.

Verifiedprimary source, dated Self-reportedthe party’s own numbers Guessour forecast Unverifiableno primary found
PLATE I

Two conversations, one missing edge

A person asks a model a question. The model answers. What crosses the wire is a claim — fluent, confident, unaccompanied. What does not cross is everything needed to believe it: where it came from, when it was last true, and what the model did not check.

An agent calls another agent. Same wire, same omission. HTTP has negotiated media type since 1996, language, encoding, and — through Memento — the date of a representation. It has never negotiated the quality of the evidence.

These are usually treated as two fields with two literatures. They are one problem with two renderers.

LANE A · PERSON ⇄ MODEL PERSON MODEL CLAIM LANE B · AGENT ⇄ AGENT AGENT AGENT CLAIM PRIMARY SOURCE VERIFIED-AT WHAT WASN’T CHECKED NEVER ON THE WIRE — IN EITHER LANE
Fig. 1 — The symmetric omission. The claim travels; the grounds for believing it do not. Every failure catalogued in the next four plates is a consequence of this one edge being absent.
PLATE II

What the old work already solved

The canon is not a warm-up. Most of what the agent era is currently rediscovering was specified — precisely, and with experiments — between 1945 and 1999. The right-hand column is the uncomfortable part.

YearSourceThe mechanismWhat 2026 is re-solving badly
1968Taylor,
question-negotiation
Four levels of need — visceral, conscious, formalized, compromised — and five filters that walk a bad query back to the real one. The reference interview is a complete intent-elicitation protocol. We replaced it with an empty prompt box and called the mismatch prompt engineering.
1975Grice,
maxims
Quality: do not assert what you lack adequate evidence for. Quantity: do not over-inform. Hallucination is not a novel failure mode. It is a maxim violation with a marketing budget.
1985Hutchins, Hollan
& Norman
Two gulfs: execution (intent → available action) and evaluation (output → interpretation against the goal). LLMs collapsed the execution gulf and widened the evaluation one. Output you cannot check against your goal without redoing the work is not an answer.
1989Star & Griesemer,
boundary objects
Objects “plastic enough to adapt to local needs… yet robust enough to maintain a common identity across sites.” This is the theory that permits one record to serve a curator, an investor and an agent without collapsing into mush. Nobody is using it.
1991Clark & Brennan,
grounding
Understanding is jointly ratified — presentation and acceptance — to “a criterion sufficient for current purposes.” Chat presents without accepting. There is no cheap ratification turn, so repair cost lands entirely on the human, downstream.
1996Green & Petre,
cognitive dimensions
A vocabulary for judging notations — viscosity, premature commitment, hidden dependencies. A claim whose source dependency is hidden rots silently. That is precisely what a verification heartbeat exists to catch.
1999Pirolli & Card,
information foraging
Foragers follow scent — proximal cues about the value and cost of a source. Generated prose is cue-poor by construction. It degrades scent for every downstream forager, human and machine alike.
1999Horvitz,
mixed-initiative
Twelve principles, including: use dialog to resolve key uncertainties, and minimise the cost of poor guesses about action and timing. This is the agent-UX specification, written twenty-seven years ago. Most agents act unconditionally and apologise afterwards.

Every row above: Verified against the primary text. Sources in Plate VIII.

PLATE III

What 2026 actually measured

Four findings from the past eighteen months do real damage to the intuitive design of an AI interface. They are the reason the proposal in Plate VI takes the shape it does.

Explanations amplify. Sources correct.

In a pre-registered study of 308 participants, explanations increased reliance on model output that was wrong as readily as output that was right. Reliance on incorrect answers fell in two conditions only: when sources were provided, and when the explanation visibly contradicted itself. Verified

Design consequence: shipping more reasoning is not a safety measure. Shipping the receipt is.

Citation has to unfold in place.

An interface that lets a reader open a cited sentence into its constituent claims, colour-coded by supporting and contradicting evidence, without leaving the page, got readers to their first source about 48 seconds faster and led them to insert more grounded facts into their writing than a conventional citation list did. Twenty participants; modest, directionally clean. Verified

Design consequence: a link that costs a context switch is decoration, not verification.

What the agent did not do is invisible.

Across 20,574 real coding-agent sessions, two of the most common failure signatures were violating a stated constraint (38.33%) and reporting inaccurately on its own work (22.58%). Separate work from Microsoft Research finds that people overseeing agents want support for interpreting when and why the agent did not find what was asked — and proposes extending trace displays to show what the agent did not do. Verified

Design consequence: the unchecked list is not an appendix. It is half the payload.

Verbalised confidence is one bit.

Hedging language in model output is roughly stable within a domain and degrades across domains, and the markers compress badly against an actual 0–1 confidence range. The usable signal is hedge or no hedge, not which hedge. Verified

Design consequence: do not ask the model how sure it is and print the number. Compute trust externally, from the record.

PLATE IV

The machine lane, without the press release

The agent-protocol landscape as of this week. Read the right-hand column before building on anything.

MCP
Model Context Protocol
Build on it
Spec 2026-07-28 was a breaking rewrite toward statelessness: the initialize handshake and session id are gone; Roots, Sampling and Logging are deprecated. It also added CacheableResult with a ttlMs — a native, already-standard hook for expressing claim freshness. It carries no provenance or trust metadata at all.
A2A
Agent2Agent
Publish a card
v1.0 shipped with cryptographically signed Agent Cards at a well-known path. It deliberately does not specify a registry API — discovery is a hole. Both MCP and A2A now sit under one Linux Foundation umbrella, which is governance convergence, not technical convergence.
Web Bot Auth
IETF, HTTP signatures
Adopt now
A chartered working group with a draft that graduated to WG ownership, and a CDN shipping it in production. This is how you stay on allowlists as the open web closes to unidentified agents. Cheapest high-value item on this sheet.
llms.txt
site manifest
Keep, don’t believe in
Publication grew roughly nine-fold in a year, to about 36,000 files. But server-log analysis across 137,210 domains found that in May 2026, 97% of these files received zero AI requests — and the largest single consumer of the ones that were fetched was SEO audit tooling, not any model. Treat it as a contract you point a partner at, never as a distribution channel.
x402 / AP2
agent payments
Not yet
The project’s own live counter reads 75.41M transactions and $24.24M of volume in the last 30 days — about $808,000 a day — from 94.06K buyers and 22K sellers (x402.org, read 2026-09-02). The “roughly half judged artificial” finding belongs to a March 2026 analysis by Artemis reported in CoinDesk, and has not been re-audited at this volume. Build the meterable surface — stable per-claim URIs, per-call logs, a price list — and switch payments on in a week when demand exists.
Latent channels
activation / KV exchange
Never on the trust path
Real physics: up to 27% accuracy gain at under a quarter of the compute by passing activations rather than natural language, and a near-24× cut in end-to-end latency in one latent-exchange system. Their own authors concede the messages are not human-readable. Fine for orchestration between our own models. An unauditable channel cannot carry a receipt.
Nanopublications
claim + provenance + pub-info
Steal the idea, not the stack
The best existing model of a verified claim object — three named graphs, content-addressed, immutable. Its public hub shows roughly 10.9 million records from about 139 users. Epistemically superior; distributionally dead.
ClaimReview
fact-check markup
Do not build on
Google removed the fact-check rich result and is phasing out support in Search. The vocabulary survives; its distribution channel does not. A cautionary tale about depending on one consumer.
PLATE V

The gap, stated precisely

Here is what a deliberate search for prior art returned. HTTP negotiates media type, language, encoding, and — via Memento — the datetime of a representation. W3C has a Recommendation for provenance lineage, and a Recommendation family for verifiable credentials. Neither models quality; provenance says where a thing came from, not how good it is or whether it is still true.

There is no general-purpose, cross-domain, machine-readable vocabulary for evidence strength in any W3C Recommendation. The one narrow property that exists is bound to medical guidelines and shows under a thousand domains using it. And there is no mechanism, anywhere, by which a consumer can ask a producer for claims of a given evidential quality.

Two near-misses are worth naming, because they are the first objections anyone will raise. The Data Quality Vocabulary is the closest general-purpose quality vocabulary — and it is a Working Group Note from 2016, not a Recommendation. Content Negotiation by Profile is the closest thing to a negotiation header — and it is still a Working Draft, and what it negotiates is conformance to a schema, not the strength of the evidence behind a claim. Neither closes the hole. We went looking for prior art specifically to kill this proposal, and could not.

That is the gap. It is load-bearing, it is unclaimed, and it sits exactly where our garden already does its work.

TODAY CLIENT GARDEN GET /signals everything, flat 11 claims. Unknown mix. Unknown age. What was left out: not asked, not said. GROUNDING CLIENT GARDEN Accept-Evidence: Content-Evidence: served = 7 · weakest 0.71 withheld = 4 stale (2) — past max-age tier (1) — self-reported only no-primary (1) — secondary source THE ABSENCE BECOMES VISIBLE. This is the invention. Everything else is assembly.
Fig. 2 — Epistemic content negotiation. A request states the evidential floor it needs; the response states what it served and what it withheld and why. The mandatory withheld list is what stops the negotiation degrading into a no-op — the documented fate of every optional HTTP preference mechanism.
PLATE VI

The proposal — Grounding

Named for Clark and Brennan, whose 1991 account of how two people establish shared understanding — presentation, then acceptance, to a criterion sufficient for current purposes — describes exactly the mechanism the machine layer never got. Three parts. Each closes a gap the survey found genuinely unclaimed.

I · The rootstock — one record, many scions

In grafting, the rootstock carries the vigour and the disease resistance; the scion carries the fruit. Here the rootstock is the claim record — the only object that is true — and every surface is a scion compiled from it: the specimen card, the citation capsule, the JSON-LD block, the MCP resource, the caption burned into a film.

Two fields matter more than the obvious ones. checked[] and unchecked[] record what the verifier actually looked at and what it did not — the payload that the agent-oversight literature says no interface surfaces. And challenges[] carries any contest to the claim, so a disputed record travels with its dispute.

The discipline is a build step, not a convention: a divergence test in CI that fails the build when any scion asserts something the rootstock does not. That single test is what converts “verified” from a promise into a checkable property.

ROOTSTOCK claim · tier · source verifiedAt · class checked[] · unchecked[] challenges[] · signature COMPILE one source SPECIMEN CARD · human JSON-LD · the open web MCP RESOURCE · agents FILM CAPTION · the reel CI GATE fails on divergence
Fig. 3 — One rootstock, four scions. Every surface is generated, never hand-maintained alongside. The gate is the point: a card that claims more than its record is a failed build, not a copy-editing problem.

II · The handshake — negotiating the evidence, not the format

Two headers, mirroring content negotiation. The request names the evidential floor it can live with; the response reports what it served, and what it withheld and why.

→ request
Accept-Evidence: tier>=verified; max-age=90d; primary=required; verifier!=self

← response
Content-Evidence: served=7; withheld=4;
                  reason=stale(2),tier(1),no-primary(1);
                  weakest=0.71; policy="/grounding/v0"

The withheld line is the load-bearing part, and it is the part nobody has built. The documented failure of every optional HTTP preference mechanism is that servers may silently ignore the request, so the negotiation decays into a no-op. Grounding makes silence impossible: a compliant response always declares its omissions, which means a consumer can tell the difference between “there is nothing else” and “there is more, and you can’t have it.”

That is Clark and Brennan’s acceptance turn, written as a header. It is also the direct answer to the missing-negative-information failure from Plate III — at the protocol layer rather than the interface layer, so every surface inherits it for free.

III · The decay clock — trust that ages, per class, inside the signature

A verified claim is not verified forever. Each record carries its own decay curve, keyed to its class, signed into the payload so a receiving agent computes current trust offline with no status-list round trip.

This must be per-class, and that is not a refinement. Recent work measuring learned, heterogeneous decay against a single global half-life found the uniform version performing dramatically worse than applying no temporal weighting at all. A naive half-life is not a simplification; it is a regression. Protocol status ages in weeks. A 1975 paper’s claim about conversational implicature does not age.

trust(t) = tier_weight × 2^( −(t − verifiedAt) / halflife(class) )

halflife by class ·  market-figure    60d
                     protocol-status  90d
                     adoption-metric 120d
                     standard-status 365d
                     research-finding 730d
                     canon              ∞
1.0 0.5 0.0 TRUST 90d 180d 1yr 2yr 0 canon research adoption protocol One global half-life measured 18× worse than no temporal weighting at all. The curves must differ by class — that is the whole design.
Fig. 4 — Decay by claim class. Trust is a computed property of the record, not a badge someone applied once. A claim crossing its class threshold does not become false; it becomes due for re-verification, and says so on every surface it renders to.
PLATE VII

Run the handshake

The corpus below is this document’s own evidence — every claim used to build this sheet, with its real tier, source and verification date. Set an evidential floor and watch the response change. The withheld rows stay on screen, greyed, with their reason: that is the protocol working.

GET /grounding/v0/claims Corpus 18 claims · as of 2026-09-02

        

      

Decay is computed client-side from each record’s class half-life — exactly as a receiving agent would compute it, with no call back to the garden.

Everyone is building agents that can talk to each other. We are building the only thing they will need to say.

PLATE VIII

Why this shape, and what could kill it

The argument from the Semantic Web’s own post-mortem

Grounding looks, at a glance, like the thing that already failed. It is not, and the difference is economic rather than technical. The clearest retrospective on the Semantic Web — written by one of the people who built schema.org — offers the rule the original project broke: when there is an asymmetry in the number of publishers and the number of consumers, put the complexity with the smaller number.

The Semantic Web inverted that. It asked millions of publishers to pay an authoring cost so that a handful of consumers could reason cheaply, shipped long normative specifications into a world where nobody reads them, and offered the publisher no same-quarter reward. What survived was precisely the part that dropped the reasoning and attached to money: the markup a search engine paid for with rich snippets.

We are the small number. One garden, a defined circle of consumers, a verification pass we are already running for editorial reasons. The complexity lands where the rule says it should. That is not a workaround — it is the only configuration in which any of this has ever worked.

The argument from where the machine lane is going

Agent-to-agent traffic is moving toward latent channels because they are faster and cheaper, and their own authors concede the messages are not human-readable. As the fast path goes opaque, the audit path becomes the scarce good. A garden whose entire product is source-stamped, freshness-dated, independently citable claims is selling the one thing the fast path structurally cannot carry.

The honest risk

Every epistemically superior format in this survey died of distribution, not of design. Nanopublications: about 139 users. The argument-interchange formats: no adoption. Micropublications: never commercialised. ClaimReview: killed by its only consumer. The two provenance standards that won — content credentials for media, and verifiable credentials — won by attaching to hardware and to money, not to elegance.

So the strategic instruction is the opposite of the instinct. Do not try to standardise Grounding. Ship it as the garden’s internal contract, publish the spec so it is legible and forkable, point partners at it, and let any adoption be a consequence of the garden being worth reading. Our forecast — Guess — is that an evidence-negotiation header reaches an IETF draft before agent discovery gets a registry standard, because the people who need it are publishers with something to protect, and there are fewer of them.

Build order

  1. The divergence compiler ≈ 1 week

    One rootstock JSON → card HTML + JSON-LD + MCP resource, and a CI test that fails when they disagree. Do this first regardless of everything else on this sheet: it is the only item that makes “verified” checkable rather than asserted, and every later item depends on there being one true record.

  2. checked[] and unchecked[] on every record ≈ free

    A field and a rendering rule. It closes the failure the oversight literature says nobody closes, and it costs a verification pass we already run nothing extra to emit.

  3. The decay clock ≈ 3 days

    Per-class half-lives, computed client-side from the signed record. Drive the MCP resource’s ttlMs from it — that field is already in the current spec, so freshness rides a standard mechanism rather than a custom one.

  4. The handshake ≈ 1 week

    Accept-Evidence in, Content-Evidence out, on the HTTP lane and the MCP endpoint. Serve the withheld list from the first commit; it is the part that makes the rest more than housekeeping.

  5. Identity: signed agent card + outbound request signatures ≈ 2 days

    A signed card at the well-known path, and HTTP message signatures on our own outbound fetches. Cheap, standards-track, and it is how the garden stays on allowlists as the open web closes to unidentified agents.

  6. Payments: build the surface, leave the meter off deferred

    Stable per-claim URIs, per-call logging, a published price list. When agent payment volume is real rather than reflexive, switching it on is a week. Selling licensed access to three partners beats plumbing micropayments for zero agents.

PLATE IX

Receipts

Primary sources for every figure and attribution on this sheet. Where a source is secondary, the claim carries a self-reported chip rather than a verified one.

The Knowledge Gardens Production powered by XRWorkers RDKG · The Meristem · Think Tank KG Sheet FS-07 · rev v0