Publish Content Telemetry 1.0 - #49
Merged
Merged
Conversation
Consultation issue #2. A SHA-256 of a client IP is a pseudonym, not an anonymous value: the IPv4 space is small enough to enumerate against a candidate digest. The field is removed from the edge and origin data profiles, the schema and the fixtures. Section 9.1 previously told emitters to "hash or anonymise identifiers where possible", which is the claim the consultation objected to. It now points emitters at asn, asn_org and country, which describe the network path rather than the client. Event data accepts additional properties, so the schemas cannot reject a withdrawn field. The conformance suite gains an application-layer rule and a negative fixture instead. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Consultation issue #6. Token counts are measured in the emitter's own tokeniser, so they are not comparable between agents and shift when a vendor revises a tokeniser. A content owner receiving grounding events from several agents could not aggregate the only volume measure the specification offered. chars_ingested counts Unicode code points in the content placed in the generation context. Two emitters counting the same text agree. tokens_ingested stays, described as supplementary, and section 5.5 now carries the same caveat for query_tokens and response_tokens. Grounding conformance asks for chars_ingested rather than tokens_ingested. Fixtures and worked examples carry both. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Consultation issue #22. Section 5.2.3 said a consumer receiving license_ref "can verify that content usage was licensed". Core neither resolves nor interprets the value, so it establishes none of that. The field is part of the emitter's claim: it records which grant the emitter says applied, not that the grant existed, covered this content, was valid at the time, or permitted the use. Also records what the field does not identify. COUNTER Metrics read license_ref as a possible carrier for the institution deriving access rights; it is not one. Where a publisher issues one grant per subscriber the value can work as a proxy inside that publisher's namespace, but nothing requires it to be typed, stable, or comparable between emitters. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Consultation issues #13 and #5, both documentation dispositions. Section 1.3 named model training as a non-goal but said nothing about fine-tuning, embeddings or index construction, so readers could not tell which side of the line those fell. New section 1.3.1 draws the line at construction against use: assembling a corpus, fine-tuning, computing embeddings and building an index are outside scope, while querying such a store during a response is an ordinary grounding event. It also states plainly that nothing here reports whether a model was trained on a work. Section 4.4 gains the limit raised by the scraper-resale path in #5. Where the intermediary in the middle does not emit, the agent's events are the only record and carry only what the agent was told. Consumers should treat the path back to the content owner as unestablished rather than infer it. The provenance fields for the same issue are PR #10 and are not touched here. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The output constructor's claim that the output artifact contains identified source content - a quotation, an excerpt, or a full copy, verbatim or near-verbatim. Reproduction and citation are sibling output-construction claims: reproduction records the material, citation records the credit; neither implies the other. Covers uncredited reproduction in unpresented output (API delivery), which no existing event can host. Char-first counts mirror the excerpt_chars/excerpt_tokens pairing on citation data. Builds on the v1 presentation semantics change and is prepared pending D1. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
A source association with no resolvable reference is not a citation. The JSON Schema now rejects a content_cited event whose content_url and content_id are both absent or null - stricter than the application-layer identifier rule that covers content events generally, because the reference is what makes the credit routable to an owner. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
A reproduction claim is only meaningful for an identified source. The schema rejects a content_reproduced event whose content_url and content_id are both absent or null, matching the enforcement added for content_cited on v1-citation-source-reference; merge that branch first so the cross-reference in section 6.6 holds. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Renames content_displayed to content_presented, distinguishes content from source_reference, and adds output/citation/presentation correlation with migration notes and multimodal fixtures. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
A source association with no resolvable reference is not a citation: content_cited events must carry a non-null content_url or content_id, enforced by the JSON Schema. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Sixth core event: the output constructor's claim that the output artifact reproduces identified source content, verbatim or near-verbatim, independent of credit (cited) and delivery (presented). Includes the source-reference schema enforcement matching content_cited. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
- Isolate the cited source-reference anyOf in both negative fixtures (they previously also omitted id/output_id and failed on those) - Update five-stage/three-model/enumeration text that predated the sixth event (1.3.1, 5.2 field table, 5.7.5, 7.3, 10.1, 10.2) - Schema-enforcement prose now names content_reproduced alongside content_cited; section 6 intro no longer claims no data field is required; terms table gains presentation - Reorder session-multi-turn events chronologically; correct stale fixture descriptions and README event names; batch grounding fixture now meets the conformance level it claims - Add invalid fixtures isolating cited output_id, reproduced id and presented presentation_type; add envelope-level parent_session_id coverage; pair chars_ingested with tokens_ingested in new fixtures - tests/README documents the ip_hash application-layer check and no longer claims batch-envelope engagement coverage Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
# Conflicts: # README.md
Complete v1 grounding provenance and fingerprint semantics
Output-side reuse is reported with content_reproduced on the landed v1-draft; grounding-time detection stays as the emitter claim. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Add `provenance` and `content_fingerprint` to `content_grounded` (v0.1)
… resolution (#23, #28, #30) New section 7.4 replaces the v0.1 click manifest with the click context: - Token pattern ct_[A-Za-z0-9_-]{8,240}, opaque, bound to exactly one presentation at mint time; per-click minting for routed surfaces. - Reserved query parameters ctx_token and ctx_iss with SHOULD-level redirect propagation mirroring section 7.2. - Resolver discovery through the issuer manifest: telemetry.ctx_resolution. - Resolution returns the engagement, the clicked content's lineage cut by content identity across turns, and an optional count-based session summary; never the raw session_id, never other owners' events. - presentation_id becomes conditional: required for agent-reported engagements, restored from issuer state for destination reports that carry ctx_token; the URL-carried presentation UUID is withdrawn. - Consumer-custody trade-off recorded (#29). Schemas, six new/updated fixtures, migration note. 74/74 conformance checks, 9/9 worked examples. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
…livery (#28, PR #38 review) Two changes from Pedro Santos's review of #38, both from #28. Contributing sources (7.4.4). Lineage by content identity alone cannot support the commerce case: a click through to a destination contributed by another owner's content resolves nothing about the contributor. The click context regains the v0.1 click manifest's role as a consent-gated contributing-source set, narrowed to the engaged presentation's turn and gated per contributing owner's recorded opt-in. Owners without an opt-in remain visible only as counts. The migration note in 12.1 now records a scoping plus the retained gate, not a removal. Owner-scoped delivery (new 7.4.5). The agent-authored click content_engaged is delivered to the engaged content's owner through 7.3 filtered views on the same terms as grounding and citation events, session_id included. The 7.4.4 session_id prohibition binds token resolution (possession of a URL-carried value), not 7.3 delivery to a verified domain owner. This survives token loss in redirect chains, removes the resolver dependency for the destination's own join, and lets a party holding both a contributor's and a destination's owner-scoped streams match on session_id for attribution. Not token distribution: only the clicked content's owner receives the event. Recorded limit renumbered to 7.4.6. Conformance suite 74/74, worked examples 9/9. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Click context: ctx_token issuance, carriage, discovery and resolution
…s_context, content_depth (#3, #42, #43, #44) Terms reference (new 5.2.4, #3). Optional event-level terms_ref, distinct from license_ref: the terms are the basis, the licence is the proof. A public URL or an opaque identifier both parties can resolve; private resolution conforms. Core carries the reference and does not resolve, validate or interpret it, and the terms it points at do not redefine core event semantics. Event-level only: a session-level default was considered and deferred because sessions legitimately span content under different terms. Occurrence and coverage (#42). Section 4.3 now states one occurrence boundary per core event: retrieval per completed fetch, grounding per distinct content item per declared scope, reproduction per source item per output element, presentation per rendering, engagement per observed action. "Qualifying occurrence" is defined, a conformance level is stated to be a capability claim rather than a coverage claim, and coverage becomes an explicit declaration (new 5.7.6): complete, sampled, aggregated or selected, over a stated relationship scope, with the rule for the last three objectively decidable at emission time and disclosed. Manifests MAY declare the modes machine-readably (8.5). Governing terms select events and coverage; they do not redefine semantics (SCOPE.md, the boundary recorded in #4). Envelope identity. manifest_ref joins the standalone event and event batch envelopes (7.1), the only envelope field naming the manifest - and so the domain - under which the emitter claims to report. SHOULD-level for standalone delivery under settlement or audit obligations. Session access context (new 5.1.3, #43). Sessions gain a data container mirroring the event-level field, closing the accidental extension point at the session root (11.1, 12.1). One core container: access_context, typed institutional identifiers (ror, saml_entity_id, isni, open vocabulary) - the institution whose access rights the session used, never an individual. Populated only where governing terms require it, placed inside the privacy model in 5.5 with intent/minimal pairing guidance. Content depth (6.1, #44). content_depth records how much of a content record a retrieval reached: metadata, abstract or full, open values, applicable from any source_role. Depth is what was reachable at retrieval, independent of what later entered a generation context. Four new fixtures. Conformance suite 78/78, worked examples 10/10. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
…eans nothing (#3) Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Terms references, coverage declarations, envelope identity, session access context and content depth
The spec body already makes v1 normative claims (ip_hash withdrawal in 9.1, the content_displayed rename in 12.1) while the header still said 0.1/Preview and every schema pinned schema_version to "0.1" under a /schema/v0.1/ $id, so a conforming v1 emitter had to declare "0.1" and the two versions were indistinguishable on the wire. - Header: Version 1.0 (release candidate draft), status release candidate in preparation, feature freeze 21 August 2026. - All four schemas: $id path v0.1 -> v1, schema_version const -> "1.0"; manifest const description now says v1 emitters MUST use "1.0". - 5.7.4 and 12.1 now state explicitly that v1 documents declare "1.0", that a v0.1 consumer rejects them under the preview rule, and that a v1 consumer rejects "0.1". - Every fixture and every inline example in SPECIFICATION.md and README.md now declares "1.0". invalid/invalid-schema-version.json keeps its deliberately wrong "2.0" (still != the new const). - validate.py / tests/README.md docstrings updated; no hardcoded versions or v0.1 paths remain in the test scripts. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The table said id was optional and generated by the server if not provided. The schema requires id on content_cited, content_reproduced and content_presented, and server-side generation is incompatible with citation_id and presentation_id, which reference an id the emitter must already hold when it constructs the referencing event. Reworded to the table's conditional style (as used by output_id and presentation_id): required for reproduced/cited/presented, optional elsewhere, emitter-assigned. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Section 5.7.3 lists data.citation_type among the Citation-emitter requirements marked schema-enforced, but the content_cited conditional required only id and output_id - unlike content_reproduced and content_presented, whose conditionals require their data objects. The content_cited conditional now requires data with citation_type, mirroring the reproduced (data.reproduction_type) and presented (data.presentation_kind/presentation_type) pattern. Two cited events carried no data and gained citation_type: "reference" (each is paired with a link presentation of the same source): the 7.1 event-batch example and tests/valid/event-batch-agent.json, plus the same event in tests/invalid/batch-missing-session-and-ctx-token.json so that fixture still passes the schema and fails at the application layer as its description intends. All other cited fixtures and examples already carried citation_type. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
- 8.3: revert 'Presentation name' to 'Display name' for operator.name. The display->presentation rename applies to the presentation event, not ordinary UI terminology; manifest.json already says Display name. - telemetry-session.json: ad_rendered description now says 'rendered', matching the field name and 5.4. - 6.5: restate the excerpt pair on the v1 chars-primary hierarchy - excerpt_chars is the portable primary measurement under 6.4's counting rule, excerpt_tokens the agent-native supplementary one - mirroring 6.4's chars_ingested/tokens_ingested wording so 6.6's 'same pairing' cross-reference holds. - 6 intro: the profiles are no longer in lifecycle order (6.5 Citation precedes 6.6 Reproduction while the lifecycle runs Reproduced then Cited), so the intro no longer claims they are; sections keep their numbers. - 5.7: the optional-signals sentence now names reproduction alongside presentation and engagement - it likewise sits outside the Retrieval/Grounding/Citation ladder as a SHOULD. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
…shapes Mutation-verified review findings, all previously undetected: - Build every validator with format_checker so format: uuid / date-time / uri assertions enforce instead of annotate; both runners hard-error at startup if the checker lacks uuid or date-time. Install line becomes pip install "jsonschema[format-nongpl]". - Require _expected_error on every invalid fixture: a substring that must appear in the actual error (first schema error message and JSON pointer, or the application-layer violation text). A fixture that fails for the wrong reason, or carries no pin, now fails the run. - Reconcile APPLICATION_LAYER_VIOLATIONS keys against invalid/: an entry with no matching file fails the run instead of silently dropping the expectation. - Apply privacy field gating (5.5) to turns wherever they appear: session documents, event batches, and standalone event envelopes, not only session events lists. - Add referential integrity checks within a session document (6.6-6.8): content_engaged.presentation_id must match a content_presented event id, and citation_id on content_presented/content_reproduced must match a content_cited event id. Standalone envelopes and batch members are exempt; the corroborating click-out flow is out of scope here. - check_examples.py: validate complete bare event objects (type + timestamp) against the TelemetryEvent definition instead of skipping them; 3 of the 7 skipped fragments are now validated. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
New invalid fixtures, each pinned by _expected_error:
- presented-missing-id, presented-missing-output-id, cited-missing-id:
the required event and output identifiers of sections 6.5 and 6.7.
- cited-missing-citation-type: the schema now requires
data.citation_type on content_cited.
- reproduction-type-invalid: reproduction_type 'paraphrase' - section
6.6 says paraphrase is not reproduction.
- presentation-kind-invalid: presentation_kind is a closed two-value
enum.
- grounded-negative-chars-ingested, reproduced-negative-chars: count
fields carry minimum 0.
- malformed-parent-session-id: format: uuid, caught only now that the
runners enforce format assertions.
- privacy-violation-{response-text,query-intent,topics,response-type,
response-mode,model-id}-at-minimal: one fixture per remaining field
forbidden at minimal privacy (section 5.5).
- presented/retrieved/engaged-missing-identifier: the section 5.7.5
identifier rule, previously only tested via content_grounded.
- engaged-presentation-id-unmatched, presented-citation-id-unmatched:
the new intra-session referential integrity checks (sections 6.7,
6.8).
New valid fixture:
- session-reproduction-no-grounding: the fifth funnel departure of
section 4.3, reproduction of memorised content with no grounding
event.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
mutation_smoke.py copies the schemas and tests/ into a temp directory, applies each known suite-weakening mutation - dropping the format checker, gutting the withdrawn-ip-hash fixture, shrinking CONTENT_EVENT_TYPES and PRIVACY_FORBIDDEN_FIELDS, pointing an engagement at an all-zeros presentation_id - and confirms validate.py fails under every one. Each of these previously went undetected. The working tree is never modified; a surviving mutation fails the script. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
…s branch The rebase onto v1-draft brought in fixtures and an inline example added by PRs #38, #41 and #45, none of which were swept by the v1 identity commit or carry the pinned violation this branch now requires. - CI installed plain `jsonschema`, so the hardened runners' format-checker guard aborted the job. The workflow now matches tests/README.md and installs `jsonschema[format-nongpl]`. - 17 fixtures and the §5.1.3 access_context example still declared schema_version 0.1. Bumped to 1.0; invalid-schema-version.json keeps its deliberately wrong value. - 8 invalid fixtures had no `_expected_error`. Each now pins its own violation, including the two application-layer provenance rules and the ctx_token pattern. Suite: 107/107 fixtures, 13/13 examples, 5/5 mutations caught. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
…ick-context gate, migration, suite hardening Prose (SPECIFICATION.md, README, tests/README) - schema_version is major.minor; 5.7.4, 8.7 and 12 state the one-schema-per-minor rule; the 1.0.0-style examples are gone and the const "1.0" stays - source_role is a MUST on content_retrieved everywhere (5.2.2 aligned with 5.7.1) - data.scope and citation_type stated as required in 6, 6.4, 6.5; 5.7.1 says "every content event" - 5.7.5 gains the field-placement, distinct-id, same-content join, one-token-one- presentation, envelope-token and manifest placement rules - 7.4.1 token minimum 16 characters and an unguessability MUST; 7.4.4 resolution gate defined on the click turn's privacy_level (minimal: engagement and lineage only; no turn fields at any level); 7.4.5 destination-owner caveat; 7.4.6 records token lifetime and requester authentication as v1 limits - 9.1 states that privacy_level gates the named turn fields only and that opaque and extension fields must not carry identity or withheld text - 12.1: the "URL-carried presentation_id" and published-v0.1 preserved_in_output paragraphs corrected; migration notes for citation_type, scope, the non-null reference rule, ip_hash, source_role and license_ref semantics added - Section 8 written in v1 tense; manifest id and endpoint are https-only and id sits at the well-known path; tokens_ingested at minimal; share and user-supplied agent_navigate engagements; Annex B.5 multi-owner catalogue (#32) Schemas - content_grounded requires data.scope; content_depth typed on retrievals; ct_ pattern {16,240}; manifest_ref format uri; envelope agent_id nullable like the session field; manifest id/endpoint https patterns Conformance suite - validate.py: source_role, field placement, duplicate ids, same-content joins, shared tokens, envelope ctx_token, root-manifest domains, agent/platform ctx_resolution; check_examples.py runs the application-layer rules too - 66 new invalid and 7 new valid fixtures: every closed enum, every format assertion, every cross-shape rule in all three document shapes, origin and index retrievals, platform manifest, all engagement/citation/presentation types, a destination engagement batch, the v0.1 wire version rejected - every schema pin is pointer-prefixed; session-minimal carries source_role; the one fixture without a description has one - mutation_smoke.py replays 59 mutations (54 new) and runs in CI Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The v1 lifecycle keeps the content_displayed -> content_presented rename and drops the separate reproduction event. Output-side reuse reporting (reproduction_type, reproduced_chars/tokens/hash, the reproduced funnel departures and per-reproduction counting) leaves core; a future version or profile can reintroduce it on implementation evidence. The migration note records that the event existed only on the pre-release draft line. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
- Discovery/outcome-layer clarification in 1.4 and the consumer vs index-emitter role separation in 7.3 (#12) - Optional manifest identifier_schemes block and co-primary URL / content_id owner resolution with defined conflict and failure behaviour (#17, #31), with fixtures and placement checks - The manifest identifies a domain, not a legal person; party-identity declarations recorded as deferred (#46) Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
…aveat - Version 1.0 / current-specification framing in the spec header, README and GOVERNANCE; the consultation section becomes a past-tense record and the checkbox list goes away - GOVERNANCE gains the decision process owed before 1.0 and loses the broken request-for-comment anchor - Cross-language citation guidance in 6.5 (#24 fallback) and the bounded-portion caveat restated in 6.6 (#25) Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
- bot_category becomes purpose: an open enum classifying the access (training, inference, search, advertising), with the vendor signal mappings moved to informative Annex C and a migration note (#7) - Generic evidence reference slot on content events (6.8): scheme, detached ref plus digest, no status escalation; the status vocabulary and trust policy stay with the evidence profile (#8) Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
v1 identity, spec corrections, and conformance-suite hardening
… link forward Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Version 1.0 replaces the v0.1 preview. The changes accepted through the 12 June - 24 July 2026 public consultation, the release-candidate hardening, and the migration notes are all in SPECIFICATION.md; section 12.1 is the complete migration record for v0.1 implementers.
Headline changes from 0.1:
content_displayedbecomescontent_presented, withpresentation_kinddistinguishing content from source references ([spec] Clarify citation/display semantics for non-chat response surfaces #14)id,output_id,citation_id,presentation_idbind citations, presentations and engagements to exact occurrences ([spec] Clarify citation/display semantics for non-chat response surfaces #14, [spec] ctx_token binds to click-out content_engaged event; share it on resolve with turn/session totals under partial visibility #28)ctx_tokenissuance, carriage, resolver discovery and gated resolution ([spec] ctx_token transport and redirect-chain propagation for click-out #23, [spec] ctx_token binds to click-out content_engaged event; share it on resolve with turn/session totals under partial visibility #28, [spec] A ctx_token alone does not identify which telemetry consumer can resolve it #30)terms_refand per-event-type coverage declarations (Add a reference to the governing terms, distinct fromlicense_ref#3, [spec] Nothing states when an event must be emitted, and that condition belongs to the licence #42)chars_ingested/excerpt_charsprimary, token counts supplementary (Token-based ingestion metrics are not portable #6)provenanceand detection-onlycontent_fingerprint(Provenance gap for third-party-sourced grounding (the scraper-resale supply path) #5, Addprovenanceandcontent_fingerprinttocontent_grounded(v0.1) #10)purposeopen enum replacesbot_category; vendor mappings move to informative Annex C (The bot_category table hardcodes current Cloudflare/Fastly classification labels #7, Proposal: telemetry for advertising and signals-driven content usage #9)identifier_schemesand co-primary URL / content_id owner resolution ([spec] Promotecontent_idscheme registration and resolution from SHOULD to a normative requirement at the Citation conformance level #17, [spec] content_id prefix resolution implemented sooner rather than later #31)datacontainer withaccess_context;content_depth;parent_session_id([spec] Session-level data container with typed access_context institutional identifiers #43, [spec] content_depth on the content_retrieved data profile #44, [spec] Where do cited/grounded/displayed fall in multi-agent (orchestrator + sub-agent) topologies? #1)ip_hashwithdrawn (ip_hashdoes not protect the client IP, and should be replaced with non-hashed fields #2); manifest scoped to domain identity, not legal persons ([spec] The manifest is said to declare identity, but nothing in it reaches a legal person #46)content_reproducedevent is not part of v1 (section 12.1)Conformance suite: 171 fixtures, 15 validated worked examples, 59 mutation checks.
🤖 Generated with Claude Code