]> VERAX TEKNOLOJI LIMITED SIRKETI

Turkey e.dogru@cedulon.com

sec Cedulon agent decision reconciliation completeness The Cedulon core document reconciles an issuer's signed Spend Receipts against an authenticated extract of a payment rail and reports, over a declared population, that no settlement lacks a receipt and no settled receipt is absent from the rail. Money is the special case that document implements. This document defines a second population on the same reconciler. A Decision Record is signed by the party that decided whether an agent may act; an Effect Extract is an authenticated list of the effects that actually occurred on a channel. An allow must be matched by exactly one effect whose content hash the record named; a refusal must be matched by none. The Decision Record claim set, the Effect Extract shape, the points at which the reconciliation departs from the spend rules, the finding codes, and one media type are defined. This revision states that the binding compares content and reference and not the order of two clocks, corrects the boundary to two adjacent documents, and records the first reading of one frozen fixture by a second, independently written reader. The text is provisional; the companion implementation carrying this profile is published.

Introduction The Cedulon core document answers one question about an agent that spends: did every settlement on the rail have a receipt behind it, and did every settled receipt reach the rail? It answers it by closing three signed objects over a declared population: the issuer's records, an authenticated extract of the counterparty system, and epoch checkpoints that total the records. The verifier holds the keys out of band and the report names the population it covered. An agent that acts without spending raises the same question with different nouns. A party decided whether the agent may reply, post, send, or call; a channel carried whatever the agent then did. Did every effect on the channel have a decision behind it? Did every allowed action occur, once, with the content that was allowed? Did anything occur that was refused? Section 19 of reserves later profiles in name only, and its Section 19.3 sketches the same completeness calculus for other consumable resources: compute, data, energy. This document is a different population on the same reconciler, decisions against effects rather than another unit of spend, and it is the first profile written out. The profile keeps the core's three roles and its verification algorithm. What changes is the record, the row, the binding between them, and the words the report uses. What does not change is measured: the companion implementation holds the spend behaviour byte for byte behind a golden file, and every rule in this document that departs from the spend rules is stated as a departure. Three documents written at the same time ask adjacent questions about the same agent, and the boundary between them is worth stating so that a reader does not take one for another. This profile reconciles signed decisions against the effects a channel carried. runs from the invocation of an action, through its classification, to an authenticated reconciliation of the effect it produced; authorization, reservation, and the provider's entry are stages on that path, not the whole of it. reconciles the exact action and its source against an observation of the effect whose source the relying party selects, so the observation is independent where that party chose an independent source and not otherwise, and it keeps missing evidence indeterminate rather than resolving it either way. covers the delivery and enforcement of a governance control on its way to one or more enforcement points. The finding this profile exists for, an effect that occurred against a refusal (), establishes that the decision and effect populations did not reconcile; it does not by itself establish whether the failure lay in control delivery, enforcement, another path, or elsewhere. Where a deployment realizes a refusal through a downstream control path, the evidence for that path is the question addresses; where no such path exists, this profile does not imply one. This document is a companion to the core document, not a revision of it. It is not an IETF working-group item. Its requirement language is provisional in the sense Section 19 of gives the structures it reserves: a direction written with the core's discipline, not a commitment, and a later revision may change it. The companion implementation carrying this profile is published ().

Terminology The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in BCP 14 when, and only when, they appear in all capitals, as shown here. Terms defined in the core document keep their meaning here: Policy Decision Point, epoch checkpoint, trust root, population, finding, warning, guarantee. The following are specific to this profile.

Decider:

The party that decides, per request, whether the agent may act. It signs Decision Records and epoch checkpoints over them. Its key is the issuer root of this profile ().

Subject:

The party on whose request the decision was taken, named in the record as an opaque identifier.

Decision Record:

A COSE_Sign1 object signed by the Decider stating one decision: allow, deny, or defer, with the request it answered, the policy it applied, and, for an allow, the reference and the content hash of the effect it allowed ().

Effect:

One thing that happened on a channel as a result of, or in the absence of, a decision: a message sent, a post made, a call placed. It is identified by a reference, classed by a short name, and bound by the SHA-256 of its content.

Channel:

The system on which effects occur and from which an Effect Extract is taken. It plays the role a rail plays in the core document.

Effect Extract:

The authenticated list of effects on one channel, for one Decider, over one window (). It plays the role a rail extract plays in the core document.

Refusal:

A Decision Record whose decision is deny or defer. A refusal expects no effect.

The population The core document's reconciler closes an issuer record against a counterparty row over a declared population. This profile fills the same three roles:

Role	Spend (core)	Decision (this document)
Issuer record	Spend Receipt	Decision Record
Counterparty row	settlement record on a rail extract	effect row on an Effect Extract
Match key	ref	ref
Content binding	amount and currency equal	allow: a row exists and effectHash is equal; refusal: no row
Record that expects no row	outcome aborted	decision deny or defer
Aggregate witness	checkpoint totals per currency	checkpoint totals per decision kind
Declared population	account, rail, window	decider, channel, window

Which population a presented document belongs to is the verifier's call, made by the profile it applies, and never the document's. A verifier applying this profile MUST read every presented record as a Decision Record and every presented extract as an Effect Extract, and MUST refuse by name a document that does not have that shape; it MUST NOT infer the population from members a body happens to carry (MUST-DP-1). The companion found the alternative wrong in both directions: a rail extract that added a member named effects was re-routed away from the spend rules it was subject to, and an Effect Extract handed to the spend rules crashed before any report existed. Under this profile a rail extract is the wrong document and is refused as one; under the spend rules an Effect Extract is refused the same way.

Decision Record A Decision Record is COSE_Sign1 with the header profile of Section 6.2 of : deterministic CBOR, alg -19 (Ed25519, ), kid mandatory and computed as the core states, an empty unprotected header refused by name if not empty, and the payload the CBOR encoding of the claim map below. The content type header parameter is application/cedulon-decision-record+cbor ().

Claim labels The labels lie in the Private Use range of the CWT Claims registry , below the block the core document uses for the Decision Token (-70301 to -70305) and the countersignature (-70401, -70402), so that no two Cedulon claim maps share a label. Every claim annotated hash carries a SHA-256 digest rendered as exactly 64 lowercase hexadecimal characters, the grammar of Section 6.1 of , and a value outside that grammar is refused by name at signing and at verification.

Label	Claim	CBOR type
-70501	decider	tstr
-70502	subject	tstr
-70503	requestHash	tstr (hash)
-70504	policyHash	tstr (hash)
-70505	inputsHash	tstr (hash) / null
-70506	decision	tstr (allow / deny / defer)
-70507	reasonCode	tstr
-70508	ref	tstr / null
-70509	effectHash	tstr (hash) / null
-70510	timestampMs	uint
-70511	nonce	tstr
-70512	prevRecordHash	tstr (hash) / null
-70513	effectClass	tstr / null

All thirteen labels are always present; a nullable claim carries CBOR null when it has no value. The thirteenth label is new in this revision; a record with twelve is refused at verification as a claim set that does not have this shape, and the companion carries no records signed under the earlier set outside its own fixtures. decider and subject are opaque identifiers chosen by the deployment. requestHash is the SHA-256 of the request the Decider evaluated. The encoding is fixed: the canonical encoding of Section 7 of when the request is a JSON document, and its UTF-8 octets when it is text. The request's fields are not fixed by this document, and a deployment MUST state what it hashes. policyHash is the SHA-256 of the canonical policy document the Decider applied. inputsHash, when not null, is the SHA-256 of whatever further context the Decider consulted, encoded the same way, so that a later reader can tell two decisions on the same request apart by what else was on the table. reasonCode is a short token the deployment defines; it is carried, not interpreted. ref is the reference under which the allowed effect will appear on the channel, and the key on which the reconciliation matches. effectHash is the SHA-256 of the content of the effect the Decider allowed, over the octets the channel will carry: for a text reply, the UTF-8 octets of the text. The Effect Extract computes the same digest over the same octets (), so equality of the two is equality of content. effectClass is the class of the effect the Decider allowed, a short name in the vocabulary the channel defines, such as a reply or a post. An Effect Extract row carries the same claim in the same vocabulary (), so equality of the two is equality of class. The class is under the Decider's signature so that what the Decider allowed cannot be read as one class by one reader and another by the next without changing what was signed; the earlier revision carried it on the row only and named the gap. timestampMs is the decision time in POSIX milliseconds. nonce identifies the record. prevRecordHash links records into the Decider's chain: it is the SHA-256 of the previous record's COSE_Sign1 octets, the same input the core's receiptHash takes on the COSE path (Section 7.1 of ), or null for the first record of a chain.

Claim rules A signer MUST refuse to sign, and a verifier MUST reject, a claim set that breaks any of the following, naming the rule in the refusal (MUST-DP-2):

• decision is one of allow, deny, defer.
• Every hash-annotated claim that is not null matches the hash grammar.
• timestampMs is a non-negative integer of magnitude at most 2^53 - 1, the uint the label table states; a CBOR decoder hands back any number, and the rule is what makes the table true.
• An allow carries a non-empty ref and a non-null effectHash. An allow that names no effect is a decision the reconciliation cannot close, and an allow that names no reference is one it cannot find.
• An allow carries a non-empty effectClass. An allow that names no class is one whose effect could be matched by a row of any class under the same reference and content, which is the substitution names.
• A refusal MAY carry an effectClass: it names the class of what was refused, and it is carried, not measured. A refusal binds to the absence of a row of any class.
• A refusal carries effectHash null. A refusal binds to the absence of an effect, never to a content hash, so a hash on a refusal would be a claim the audit cannot measure and a second reading of whether the effect occurred. A refusal MAY carry a ref: it names what was refused, and an effect appearing under that reference is the worst finding this profile has ().

The verifier MUST apply these rules itself, on the claim map it decoded from the signed payload, and MUST NOT rely on the signer having applied them (MUST-DP-3). The Decider is the party under audit. A Decider that signed a well-formed COSE_Sign1 over a claim map that skips a rule has produced an object whose signature verifies, and a verifier that checked only the signature and the equality of the decoded map with the presented claims would attest it. The companion implementation did exactly that until it was measured: an allow with no reference, signed below the signer's own rules under the pinned decider key, verified true, was attested, was counted as unmatched, and the audit still said the books balanced. The rules now run at both ends.

Presentation and confusion A Decision Record is presented as Section 6.3 of states for the core's COSE objects: the signed octets, the decoded claim set, and the Decider's public key as a SubjectPublicKeyInfo PEM beside them. The carried key is not an identity source. Under a pinned decider key a record that verifies under the pin while carrying another key is reported as carried-key-mismatch, a warning, and stays attested; with no pin held the signature check that runs against the carried key says the record is internally consistent and nothing about who signed it. A Decision Record is not a Decision Token. The core's Decision Token (Section 8 of ) is the portable encoding of a PDP allow, carried by the party that will spend; a Decision Record is the Decider's own log of what it decided, kept for audit, and it exists for refusals as well. The two carry different content types and different claim maps. A verifier MUST reject a Decision Record whose content type is not application/cedulon-decision-record+cbor, and MUST reject a token presented as a record or a record presented as a token, on the content type, before the signature is checked and before any claim is read (MUST-DP-4).

The Decider's chain and checkpoints Decision Records chain on prevRecordHash the way Spend Receipts chain on prevReceiptHash, and the Decider signs epoch checkpoints over them with the checkpoint claim set of Section 11.1 of unchanged: receiptCount is the number of records in the window, chainHeadHash is the SHA-256 of the last record's COSE_Sign1 octets, and totals is a map from the three decision kinds to decimal counts, {"allow": n, "deny": n, "defer": n}, each rendered as a text string as the core renders its currency totals. A verifier compares the totals it computes over the attested records in the window against the signed map, and a difference is checkpoint-total-mismatch as in the core. Two records that claim the same position in a chain cannot both link to it: the second record's prevRecordHash must be the first's hash, so a Decider that signs two decisions under one nonce, or presents one record twice to the same reader, breaks its own chain and the walk names the break. A verifier MUST walk the chain over every presented record that carries the pinned decider key, not only over the records that verified, so that a record that claims the pin and fails the rules is named by the walk rather than dropped from the population without a word (MUST-DP-5). What the walk establishes is bounded by what one reader holds. A Decider can sign two successors to the same predecessor and show one branch to one reader and the other branch to another; each reader walks a linear chain that verifies, and neither walk names a break. The earlier revision called the chain the equivocation control of this profile, which overstated it and contradicted the core: Section 11 of states that the presented chain alone cannot satisfy the equivocation requirement, because its epochs are consecutive by construction, and that the comparison which reaches a fork is between the presented checkpoints and the copies a witness recorded (MUST-T11-3). That rule applies to this profile unchanged. The Decider's epoch checkpoints are the Signed Statements the witness records, the verifier compares the witness's copies against the presented chain, and two verified checkpoints for one epoch with different hashes are equivocation. Where no witness was consulted, the chain controls equivocation within the population one reader holds and nothing beyond it, and a report MUST NOT present the chain as settling more than that (the core's MUST-T11-9, with the nouns renamed).

Effect Extract Profile A verifier checks completeness against an Effect Extract, not against the Decider's own records alone. The extract is the channel's account of what happened, obtained independently of the Decider, and the profile is only as strong as that independence ().

Body and row schema The extract body is one JSON document with exactly this shape:

Member	JSON type
deciderId	string (non-empty)
channelId	string (non-empty)
windowStartMs	number (POSIX milliseconds, a safe integer)
windowEndMs	number (POSIX milliseconds, a safe integer, greater than windowStartMs)
effects	array of effect rows

Each effect row is a JSON object with exactly these members:

Member	JSON type
ref	string (non-empty; the reference the Decision Record named)
effectHash	string (SHA-256 of the effect's content, 64 lowercase hex)
effectClass	string (non-empty; a short class name the channel defines, such as a reply or a post, in the vocabulary the Decision Record's effectClass uses)
timestampMs	number (POSIX milliseconds, a safe integer, inside the window)
actor	string (optional; the party the effect reached)

These member names are normative. The body and its rows follow the core's rail extract (Section 9 of ) in every rule that document states for a JSON body: the text is read for a repeated member name before it is parsed and refused as json-duplicate-key; integers are safe integers; the window is half-open and MUST end after it starts; a missing member, a wrong type, an empty identifier, or a hash outside the grammar is refused by name at both ends, by the signer before it signs and by the verifier before it checks a signature. The profile departs from the rail extract at two points, and a reader who knows the core should note both (MUST-DP-6):

• A member this document does not name is refused, on the body and on a row. The core lets a rail add members of its own because a rail is a system the profile does not control; an Effect Extract is produced by a process the deployment does control (), and the companion measured what a free member can do to a population (). A later revision may open this once a channel that needs its own members is measured.
• A row whose timestampMs falls outside [windowStartMs, windowEndMs) makes the whole extract malformed, refused as effect-outside-window before any signature is checked. The core accepts such a rail extract and names the row (extract-scope-mismatch). Here the extract is the deployment's own document and a window it did not keep is a document it did not produce correctly: a signer applying this schema never produces such an extract, and a verifier refuses one that is presented as a document, whatever else it may also name about its rows. The trade is stated so it can be reversed: a single row out of place fails the whole window closed.

effectHash on a row is computed by the extract's signer over the same octets a Decider hashes for its effectHash claim: the content as the channel carried it. A deployment MUST state those octets once for both sides; the companion's example channel hashes the UTF-8 octets of the message text.

Authentication and scope The extract is signed the way a rail extract is signed: Ed25519 over the UTF-8 octets of the encoding of the body, with the signature as base64 and the signer's public key as a SubjectPublicKeyInfo PEM beside the body, neither inside the signed octets. It is a JSON document with a detached signature, not a COSE object, and like the rail extract it has no media type. The earlier revision gave as the reason that the core registers names only for objects whose content type is checked inside a protected header; that test decides what a name must be bound to, not whether a representation needs one, and it is withdrawn as the reason. The reason is : which population a presented document belongs to is the verifier's call, made by the profile it applies and by the decider, channel, and window it declares (MUST-DP-1, MUST-DP-7), and never the document's. That declaration is the typed outer context a media type would otherwise supply. A name on the extract would be a self-description the verifier is told not to select on, and this document does not register one that its own rule forbids relying on. The day an extract is itself wrapped as a Signed Statement and needs a content type in a protected header, an application/cedulon-effect-extract+json registration is the name to make, without changing the verifier's rule; the extract is not recorded with a witness today. Section 9.3 of applies unchanged: a signature proves internal consistency and not origin; the verifier MUST hold the extract signer's key out of band and MUST compare keys as SubjectPublicKeyInfo DER; with no key held the guarantee is conditional and unauthenticated-extract is reported; with a key held, an extract that does not verify under it is extract-key-mismatch and its rows are not reconciled (settlement-comparison-skipped, the core's name for the same condition). The extract is scoped to one Decider, one channel, and one window, and Section 9.4 of applies with the nouns renamed: a verifier that knows which Decider, channel, and window it audits MUST check the extract against them and MUST fail closed on a mismatch (extract-scope-mismatch); one that has not stated the window MUST report unstated-audit-window, one that has not stated the Decider or the channel MUST report unstated-audit-scope, and in either case the guarantee is conditional. The strongest line this profile can print, a balanced audit under an unconditional guarantee, is true of one Decider, on one channel, over one window, and a report that carries it MUST also carry those three (MUST-DP-7).

Reconciliation The verification algorithm of Section 11.4 of runs unchanged over this population: establish the subject, verify the extract, check scope, resolve records against the decider root, walk the chain, index both sides by ref, match, decode and walk the checkpoints, consult the witness if one is supplied, and decide. This section states only what the algorithm reads differently.

What binds A Decision Record expects a row when its decision is allow, and expects none when it is a refusal. For a ref that appears once on each side:

• an allow and a row bind when the row's effectHash equals the record's effectHash and the row's effectClass equals the record's effectClass; a difference in the hash is effect-mismatch (the content that occurred is not the content that was allowed), and a difference in the class with the hash equal is effect-class-mismatch (the content that was allowed occurred as something else). The hash is compared first; a row that differs in both is reported for its content;
• an allow with no row is decision-without-effect;
• a row with no record is effect-without-decision;
• a row whose ref a refusal names is effect-against-refusal.

The last is the finding this profile exists for. A spend audit has no row that should not be there in the same sense: an aborted receipt that still carries its reference and a settlement under that reference is reported by the core as a settlement without a receipt, and it never told the two cases apart. Here a refusal that was followed by the effect it refused is a different fact from an effect nobody decided on, and it has its own name. A ref that appears more than once on a side is duplicate-ref as in the core, and the repeating reference is then reconciled by count rather than by amount: there is nothing to sum. More rows than records under one reference is effect-without-decision; more records than rows is decision-without-effect. There is no amount, no currency, no manifest, no terms, and no counterparty axis on this profile. The core's counterparty-unbound scope record is not emitted: effectHash binds the content of the effect itself, which is more than a payee name ever bound on spend, and actor on a row is carried for the reader, not measured. The core's boundary rule applies unchanged: an unmatched item inside the clock-skew allowance of a window edge is boundary-deferred, and a closing-edge allow whose ref the following extract names is carried, not a finding. The Effect Extract declares no allowance of its own (); the verifier applies the core's default, which the companion holds at five minutes. The binding does not order the two clocks. A row's timestampMs is checked against the window it sits in and not against the timestampMs of the record it answers, so a row dated before the record it binds to binds as if it had followed it: the profile compares content and reference, not sequence. The core's spend binding is the same, and neither document has claimed otherwise; the gap is named here because a reader of the companion measured it and found it unstated. A deployment that needs the order can compare the two stamps within the same allowance; this revision names no finding for that comparison, and a later revision may.

Conservation With |R| the in-scope Decision Records and |E| the effect rows, the identities the core report publishes hold with the words changed: A report under this profile MUST publish these counts and MUST name the population they were computed over (MUST-DP-8), for the reason the core gives: a report whose counts do not close is a report that lost a record somewhere, and a reader is entitled to see that without re-running the audit. The report publishes refusals as one count, the core's aborted; the split of that count into deny and defer is the checkpoint's totals (), not a counter of the report.

Finding codes The identifiers below are for diagnostic output and are not an interoperability surface, as Section 11.5 of states for the core's codes. Five are new to this profile:

Code	Effect	Meaning
decision-without-effect	audit fails	An allow names a reference under which no effect occurred, or a reference had more records than rows
effect-without-decision	audit fails	An effect occurred under a reference no Decision Record names, or a reference had more rows than records
effect-against-refusal	audit fails	An effect occurred under a reference a refusal names
effect-mismatch	audit fails	The effect that occurred does not carry the content hash the allow named
effect-class-mismatch	audit fails	The effect that occurred carries the content hash the allow named and a class the allow did not

The remaining codes a report under this profile can carry are the core's, with the same effect on the verdict and the guarantee: duplicate-ref, boundary-deferred, receipt-chain-break for a break in the Decider's chain (signature, rule, or link), checkpoint-total-mismatch, checkpoint-head-mismatch, window-coverage, equivocation, unauthenticated-extract, extract-key-mismatch, extract-scope-mismatch, extract-settlement-mismatch (a caller-supplied row list that disagrees with the extract), settlement-comparison-skipped, trust-key-unreadable, unauthenticated-issuer, issuer-key-mismatch, carried-key-mismatch, unstated-audit-window, unstated-audit-scope, the witness codes, and malformed-policy-hash; the other hash claims of a Decision Record are refused at verification () and reach the chain walk rather than a malformed-hash code. Two code names carry a spend noun onto this profile (receipt-chain-break, settlement-comparison-skipped); they are kept so that one catalogue serves both populations, and a later revision may add decision-side aliases. The codes the core defines for a Trade Manifest, a payee countersignature, a beneficiary, or a counterparty are not reachable on this profile. The sentences a report prints beside those codes are another matter. An operator reading a decision report SHOULD NOT have to translate "settlement" as "effect" or "receipt" as "decision record"; an implementation SHOULD print the sentence in the population's own words, and the companion holds that under a test that runs every conformance case and refuses a spend noun in any decision sentence. Counter names in a returned structure are diagnostic and MAY keep the core's names.

Trust roots This profile has two roots, filling the core's issuer root and rail root (Sections 10.1 and 9.3 of ):

The decider root:

The key under which Decision Records and their checkpoints are attested. Everything Section 10.1 of the core states for the issuer root applies: a pinned key attests by signature, a carried key is not an identity, a record under another key is issuer-key-mismatch and covers nothing, and with no pin unauthenticated-issuer makes the guarantee conditional.

The effect-extract root:

The key under which the Effect Extract is attested. Everything Section 9.3 of the core states for the rail key applies.

The core's payee, witness, decision-token, and manifest roots are not used by this profile, except that a transparency witness MAY hold the Decider's checkpoints exactly as it holds a Receipt Issuer's, with the witness root and codes of the core unchanged. What the two roots do not cover is the relation between them. The profile's claim is only as strong as the independence of the party that signs the Effect Extract from the party that signs the Decision Records. Where the channel operator signs an export of its own log, the extract root is that operator's key and the independence is the operator's. Where the channel operator signs nothing, which is the common case for a messaging platform, the extract is produced by a capture process the deployment runs, and the deployment is stating, by pinning that process's key, that the process is not the Decider and cannot be told what to omit. The rule is not that the two roots be independent; it is that the deployment say which case it is in, and that the guarantee fall when independence is absent or unknown. A deployment MUST state which of the two it has, and a verifier MUST treat the guarantee as conditional where the extract root and the decider root are, or may be, the same party (MUST-DP-9). The companion cannot measure that from the keys alone; two keys can be held by one hand.

Security Considerations The core's threat analysis (Section 16 of ) applies where the nouns carry over: forgery and repudiation of a signed record (T4), key leakage (T7), and suppression of checkpoints (T11) are the same threats against a Decider that they are against a Receipt Issuer, and the controls are the same. The threats below are the ones this population adds or sharpens.

D1: An effect occurs against a refusal The agent, or something acting through its channel access, does what the Decider refused. This is the threat the profile exists for. A Decision Record for the refusal, with the reference it refused, and an authenticated extract that carries an effect under that reference make the event effect-against-refusal, and the audit fails. A refusal that carried no reference cannot support this finding; the effect is then effect-without-decision, which also fails the audit but does not say it was refused. A Decider SHOULD carry the reference on a refusal whenever the channel assigns one before the decision. The shape of the threat is not particular to this profile. A reader who mapped this document's objects onto a second evidence format, with a signed decision predicate and a signed outcome predicate of its own, measured that format's verifier on 4 September 2026: a decision carrying a refusal and an outcome recording execution, citing that decision by content root, verified clean with no warning, because the verdict is not read on the outcome path . That is D1 as a verifying pair rather than a finding, recorded by the reader as a measurement of that artefact and cited here as one.

D2: Effects without decisions Something acts on the channel that never asked. Every such effect is effect-without-decision. The control is the extract's completeness, which is the extract root's independence (); a capture process the Decider controls can leave the effect out.

D3: Substitution of content The Decider allows one content and the channel carries another. The allow's effectHash and the row's effectHash are computed over the same octets, and a difference is effect-mismatch. The control fails open if the two sides hash different octets, which is why a deployment MUST state the octets once for both ().

D4: The Decider signs below its own rules A Decider produces a well-formed signature over a claim map that breaks a rule this profile states: an allow with no reference or no content hash, a refusal with a content hash, a hash outside the grammar. Every such record is a record the reconciliation cannot close or would close wrongly. The verifier applies the rules on the decoded payload (MUST-DP-3) and walks the chain over every record that claims the pin (MUST-DP-5), so the record is refused and named rather than attested or dropped.

D5: Equivocation on the record chain The Decider signs two decisions for one request, or two successors to one predecessor, and offers each to a different reader. Within one reader's population the chain (MUST-DP-5) makes the second record unlinkable: it names the same predecessor as the first, or none, and the walk reports the break. Across readers the chain sees nothing: each holds a linear chain that verifies. The control for that case is the core's witness (): the Decider's epoch checkpoints are recorded with a witness the verifier has pinned, a Transparency Service being one, and the copies the witness holds are compared against the presented chain. A verifier that consulted no witness has not measured cross-reader equivocation. The guarantee its report prints is the core's, which is defined over the extract, the pins, and the window and does not cover suppression or equivocation beyond the presented chain when no witness was consulted (Section 11 of , MUST-T11-9); measured on the companion, such a report prints unconditional with no warning and no finding, and is silent about the witness it was not given. The silence is this revision's known gap: the claim is narrowed here, in the text, and a report line that names an unconsulted witness is a change to the reconciler both populations share, not made in this revision. A deployment that needs the property records its checkpoints with a witness and gives its readers the witness key.

D6: The class of the effect is substituted A row of a different class under the same reference and the same content hash: the same text allowed as a reply and carried as a post. The earlier revision named this as a gap, because the record carried no claim for the class and a row of any class matched. The record now carries effectClass under the Decider's signature (), an allow without one is refused (), and a row whose class differs from the allow's with the content hash equal is effect-class-mismatch (). What remains open is the vocabulary: the class names are the channel's, this document does not fix them, and a deployment SHOULD state them beside its statement of what it hashes so that two readers compare the same words.

D7: Silent defaults in capture The process that turns a channel's log into Decision Records or Effect Extract rows fills a missing value with a default, and the default hashes to something. An allow with no stated content that is hashed as the empty string produces a record that will match an empty effect and mismatch every real one, with no finding that says the content was never stated. Such a process MUST refuse the line by name rather than fill it (MUST-DP-10). The companion's example adapter did fill it until it was measured, and refuses it now.

D8: The capture process is the Decider The party that produces the Effect Extract is, or answers to, the party that signed the Decision Records. Every finding in D1 and D2 can then be made to disappear by omission, and no signature check detects it. This is the independence statement of (MUST-DP-9), and it is a deployment fact the profile can name but not prove. A verifier that holds both roots and cannot state their independence has a conditional result, and MUST say so.

Privacy Considerations A Decision Record carries no request content and no effect content: hashes of both, an opaque subject identifier, and a reason code. The Effect Extract carries a reference, a class, a content hash, a time, and optionally the identifier of the party the effect reached. The core's Privacy Considerations (Section 15 of ) apply to what a transparency witness is given. Two points are specific to this population. A content hash over a short text is a fingerprint of that text: a reader who can guess the message can confirm the guess. This revision hashes the plain content octets, as the companion does, so that the two sides need no shared secret to agree; a keyed or salted construction that would defeat the guess is a claim-set change and is not defined here. A deployment whose effects are short and guessable SHOULD treat the extract and the records as confidential to the audit. The subject and actor identifiers are opaque to the profile but need not be opaque to a reader; a deployment SHOULD pseudonymize them before either object leaves its control.

IANA Considerations This document requests the registration of one media type in the "Media Types" registry , in the standards tree, carrying the +cbor structured syntax suffix that registers, on the terms Section 17 of states for that document's six: it names the one COSE_Sign1 object this document defines and is checked inside that object's protected header, which is why the name cannot stay unregistered while that check stands. The Effect Extract is a JSON document with a detached signature and has no media type; states why, and names the registration that would become necessary if that changed. Registration in the standards tree requires IETF approval; until then, an implementation outside a closed deployment should treat the name as a placeholder that a registration may change. The provisional registration procedure of Section 5.2.1 is available to an Internet-Draft, and a provisional entry, if one is made, is superseded by the registration this section requests. The claim labels this document assigns, -70501 through -70513 (), lie in the Private Use range of the "CBOR Web Token (CWT) Claims" registry , integer values less than -65536, and this document requests no assignment for them. No other IANA action is requested.

application/cedulon-decision-record+cbor

Type name:

application

Subtype name:

cedulon-decision-record+cbor

Required parameters:

N/A

Optional parameters:

N/A

Encoding considerations:

binary. A COSE_Sign1 structure in deterministic CBOR , untagged, as profiled in and in Section 6 of .

Security considerations:

See of this document. The object is signed by the party under audit; its evidentiary weight depends on the verifier holding the decider key out of band () and on the verifier applying the claim rules of itself, never on a key the object carries or on the signer's word that the rules were applied.

Interoperability considerations:

The claim set is a CBOR map with the labels and types stated in , encoded per Section 4.2.1. A decoder refuses a duplicate key, an input beyond its stated bounds, and a non-empty unprotected header by name rather than accepting it, as Section 6 of requires of every Cedulon object. A Decision Record and a Decision Token are distinct objects with distinct content types and are never accepted for one another.

Published specification:

This document, .

Applications that use this media type:

Policy decision points and other deciders that log the decisions they take about an agent's actions, and auditors that reconcile those logs against the channels the actions occurred on.

Fragment identifier considerations:

N/A

Additional information:

Deprecated alias names for this type: N/A. Magic number(s): N/A. File extension(s): N/A. Macintosh file type code(s): N/A.

Person and email address to contact for further information:

Emek Can Dogru, e.dogru@cedulon.com

Intended usage:

COMMON

Restrictions on usage:

N/A

Author:

Emek Can Dogru

Change controller:

IETF

Implementation Status This section is to be removed before publishing as an RFC. RFC 7942 note.

Implementation:

The profile is carried by the core document's companion implementation at https://github.com/dogrucanemek-alt/cedulon, on the same reconciler that implements the core, selected by a profile object rather than by a second code path. As of the commit this revision was written against, the tree carries the Decision Record and Effect Extract objects, the profile, twenty conformance cases covering the rules and departures this document states, and four offline fixtures for one example channel, a direct-message reply log. The two cases added with -01 were red before the claim was added: a row of a different class under a matching hash matched, and an allow signed without a class verified. This revision changes text and adds no case; the ordering of the two clocks () is stated, not enforced. The spend behaviour of the same reconciler is held byte for byte by a golden file of fifteen cases generated from the source before the profile seam was added.

Maturity:

Published. The code carrying -01's claim set, the class under the signature and the fifth finding, was released as the companion's 0.13.0 on 4 September 2026, nine packages with provenance, after -01 was posted. Before merge the branch was read twice: by the author's own gate, re-running it from a second worktree, and by an outside model reading the whole diff and barred from changing it. The two readings found four defects in the first cut of this profile, each recorded in the companion's review log: a verifier that did not apply the signer's claim rules (D4), a spend-side crash on the wrong document (), a spend warning leaking onto decision reports, and a document that counted its own cases wrong. All four were closed with a test that was red before the fix. A third pass moved the report vocabulary onto the profile.

Not measured: a live channel log. The example adapter maps a proposed line format for a direct-message bridge; the bridge's actual field names were not read when this revision was written, and the adapter is written so that only its two line-mapping functions should move when they are. No independent implementation of this profile is known to the author.

Read by a second reader: one frozen fixture, a leaked refusal in the example channel's line format, was read by this implementation and, on 4 September 2026, by an adapter written by the author of for that document's verifier, with the two results recorded side by side the next day. This reader reported effect-against-refusal on the leaked row; the other reported the case as reconciled, divergent, and not valid. The three fixture digests, both result digests, and the two verdicts are in the companion's external review log. The limits are the other reader's, and they are kept with the record: one pinned raw fixture read through separately owned adapters is not native-format interoperability; the test keys are public and the timestamps were written after the fact, so the run establishes neither identity nor temporal precommitment; and neither reader says where the refusal failed.

Reproduced: the companion's suite at the commit that posted -00 was run unmodified by two readers on their own machines, one on the day of posting and one after 0.13.0 shipped. The second run had one red, a test that asked the package registry what it served that day from inside a pinned commit; the suite now reports that case as the world having moved rather than as a failure, and the change is the reader's. The first reader also put three probes to the profile: an effect dated before its decision, which is what now states; one content hash under two allows, which the binding keeps apart by reference as written; and a record altered after signing, which the chain walk already names as a break. The -00 text was read by the author of and the day it was posted; the four items that reading raised were the changes of -01, and the two boundary sentences of this revision are that reader's as well.

References Normative References In many standards track documents several words are used to signify the requirements in the specification. These words are often capitalized. This document defines these words as they should be interpreted in IETF documents. This document specifies an Internet Best Current Practices for the Internet Community, and requests discussion and suggestions for improvements. Federal Information Processing Standard, FIPS This document defines procedures for the specification and registration of media types for use in HTTP, MIME, and other Internet protocols. This memo documents an Internet Best Current Practice. This document describes elliptic curve signature scheme Edwards-curve Digital Signature Algorithm (EdDSA). The algorithm is instantiated with recommended parameters for the edwards25519 and edwards448 curves. An example implementation and test vectors are provided. RFC 2119 specifies common key words that may be used in protocol specifications. This document aims to reduce the ambiguity by clarifying that only UPPERCASE usage of the key words have the defined special meanings. CBOR Web Token (CWT) is a compact means of representing claims to be transferred between two parties. The claims in a CWT are encoded in the Concise Binary Object Representation (CBOR), and CBOR Object Signing and Encryption (COSE) is used for added application-layer security protection. A claim is a piece of information asserted about a subject and is represented as a name/value pair consisting of a claim name and a claim value. CWT is derived from JSON Web Token (JWT) but uses CBOR rather than JSON. Cryptographic operations like hashing and signing need the data to be expressed in an invariant format so that the operations are reliably repeatable. One way to address this is to create a canonical representation of the data. Canonicalization also permits data to be exchanged in its original form on the "wire" while cryptographic operations performed on the canonicalized counterpart of the data in the producer and consumer endpoints generate consistent results. This document describes the JSON Canonicalization Scheme (JCS). This specification defines how to create a canonical representation of JSON data by building on the strict serialization methods for JSON primitives defined by ECMAScript, constraining JSON data to the Internet JSON (I-JSON) subset, and by using deterministic property sorting. The Concise Binary Object Representation (CBOR) is a data format whose design goals include the possibility of extremely small code size, fairly small message size, and extensibility without the need for version negotiation. These design goals make it different from earlier binary serializations such as ASN.1 and MessagePack. This document obsoletes RFC 7049, providing editorial improvements, new details, and errata fixes while keeping full compatibility with the interchange format of RFC 7049. It does not create a new version of the format. Concise Binary Object Representation (CBOR) is a data format designed for small code size and small message size. There is a need to be able to define basic security services for this data format. This document defines the CBOR Object Signing and Encryption (COSE) protocol. This specification describes how to create and process signatures, message authentication codes, and encryption using CBOR for serialization. This specification additionally describes how to represent cryptographic keys using CBOR. This document, along with RFC 9053, obsoletes RFC 8152. This specification refers to cryptographic algorithm identifiers that fully specify the cryptographic operations to be performed, including any curve, key derivation function (KDF), and hash functions, as being "fully specified". It refers to cryptographic algorithm identifiers that require additional information beyond the algorithm identifier to determine the cryptographic operations to be performed as being "polymorphic". This specification creates fully-specified algorithm identifiers for registered JSON Object Signing and Encryption (JOSE) and CBOR Object Signing and Encryption (COSE) polymorphic algorithm identifiers, enabling applications to use only fully-specified algorithm identifiers. It deprecates those polymorphic algorithm identifiers. This specification updates RFCs 7518, 8037, and 9053. It deprecates polymorphic algorithms defined by RFCs 8037 and 9053 and provides fully-specified replacements for them. It adds to the instructions to designated experts in RFCs 7518 and 9053. Informative References This document describes a simple process that allows authors of Internet-Drafts to record the status of known implementations by including an Implementation Status section. This will allow reviewers and working groups to assign due consideration to documents that have the benefit of running code, which may serve as evidence of valuable experimentation and feedback that have made the implemented protocols more mature. This process is not mandatory. Authors of Internet-Drafts are encouraged to consider using the process for their documents, and working groups are invited to think about applying the process to all of their protocol specifications. This document obsoletes RFC 6982, advancing it to a Best Current Practice. Traceability in supply chains is a growing security concern. While Verifiable Data Structures (VDSs) have addressed specific issues, such as equivocation over digital certificates, they lack a universal architecture for all supply chains. This document defines such an architecture for single-issuer signed statement transparency. It ensures extensibility and interoperability between different transparency services as well as compliance with various auditing procedures and regulatory requirements.

Changes from -01 This section is to be removed before publishing as an RFC.

• The binding does not order the two clocks: a row dated before the record it answers binds as if it had followed it, on this profile and on spend alike. Stated in , with no finding named; the Effect Extract's lack of a clock-skew member and the default the verifier applies are stated beside it.
• Two boundary sentences corrected: runs from invocation to an authenticated reconciliation and is not only what precedes an effect; the observation in is independent where the relying party chose an independent source, not by construction.
• records the companion as published at 0.13.0, one fixture read by a second reader with the limits of that run, and two reproductions of the suite.
• D1 cites a measurement of the same threat on a second evidence format, whose verifier accepts a refusal beside an executed outcome.
• Two sentences tightened after a reader's mapping: the request encoding is fixed and the request's fields are not (); MUST-DP-9 is the statement and the downgrade, not a demand for independence ().

Changes from -00

• effectClass is a claim on the Decision Record (-70513), required on an allow; a row of a different class under a matching hash is effect-class-mismatch. D6 is closed rather than named. Claim-set change; a twelve-label record is refused.
• The Decider's chain is no longer called the equivocation control of the profile. It names a break within one reader's population; a fork shown to two readers is the core's witness comparison, applied unchanged, and a report over an unwitnessed chain claims no more (, D5).
• The reason the Effect Extract has no media type is restated: the population is the verifier's declaration, not the document's (). The protected-header test is withdrawn as the reason.
• The boundary to , , and is stated in the introduction.

Acknowledgments This profile is the first concrete instance of the generalization the core document reserves. The refusal that was followed by the effect it refused, as a finding with its own name, came out of watching a messaging assistant's decision log beside the channel's sent log and finding that the spend vocabulary had no word for it. Iman Schrock read -00 the day it was posted and raised the four points this revision answers: the fork two readers cannot see, the class that was not under the signature, the media-type reason that did not hold, and the three documents that should be cited. The sentence on what an effect against a refusal does and does not establish is Ali Toygar Abak's, in his words. Pablo Etcheverry ran the suite at the posted -00 commit and asked three questions of it; the unordered clocks of are his finding. Nicholas Templeman ran the same commit after 0.13.0 shipped and showed that a claim about the present, frozen into a pinned commit, cannot tell a reproducer whether the code or the calendar failed. Konrad Gruszka mapped the Decision Record and Effect Extract field by field onto his own format and measured the D1 pair on it. Iman Schrock corrected the two boundary sentences and read the frozen fixture with an adapter written for that document's verifier.