# Vercy AI instruction - YAML 1.2 (JSON-compatible) { "vercy": "1.0-draft", "publication": { "status": "published", "adjudicationStatus": "reviewable-draft", "publishableCanonical": false, "generatedAt": "2026-10-21T17:43:25Z", "synthesisSha256": "83ce71510701d6aa26b06f162c6e81df19b7ebe478b6cf4c7713113dbee51eff", "providerMode": "single-provider-waiver", "providers": [ "Codex" ], "waivedProviders": [ "Claude", "Grok" ] }, "metaModel": { "id": "WM-MEM-015", "registryId": "vr.wm-mem-015", "name": "Retrieval Passage", "version": "0.1.0", "previousVersions": [], "entryKind": "entity", "family": "World Models", "category": "Information and virtual systems", "industry": [ "Cross-industry" ], "domain": [ "INF.MEM.PSG" ], "tags": [ "retrieval", "passage", "inf.mem.psg" ], "status": "published" }, "canonicalUrl": "https://ver.cy/models/wm-mem-015-retrieval-passage/", "sourceUrl": "https://github.com/ver-cy/world-models/tree/feat/mega-model-registry/research/runs/wm-mem-015", "model": { "registry_id": "vr.wm-mem-015", "model_id": "WM-MEM-015", "name": "Retrieval Passage", "entry_kind": "entity", "purpose": "Describe one identifiable unit offered for retrieval with explicit source lineage, representation and governed availability.", "scope_statement": "One passage entity with versioned content or a reference-only payload, anchored to one source representation per revision. It may be a bounded segment or an explicitly declared whole-source unit. Material text, source-state or segmentation changes create a successor revision or passage, never a silent remap. A retrieval occurrence only references it.", "in_scope": [ "Passage identity, class, source reference and reproducible selector", "Representation properties, extraction lineage and measured quality", "Optional index bindings and retrieval occurrence references", "Passage availability, policy bindings, provenance and local maintenance contracts" ], "out_of_scope": [ "Source-document authorship, editing and master lifecycle", "Corpus management, chunking execution, indexing, embedding generation and ranking algorithms", "Search-session or query execution and global relevance evaluation", "Answer generation, truth adjudication and claim-to-evidence entailment", "Context-policy authoring, authorization enforcement and destructive external actions", "Multi-source synthesized answers or arbitrary multimodal segments without a separate adoption profile" ], "boundary_notes": [ { "neighbor": "Source document and representation master", "distinction": "The passage carries a mandatory source reference and revision evidence; it does not own or overwrite the document. Missing source evidence leaves a draft unverified and ineligible for serving.", "source_refs": [ "SRC-001", "SRC-002" ] }, { "neighbor": "WM-MEM-014 Retrieval Index", "distinction": "Registry parent_ids is a candidate taxonomy clue, not inheritance of corpus or index operations. Use optional index references; one passage may have zero or several index memberships and remain identifiable after removal.", "source_refs": [ "SRC-003", "SRC-007" ] }, { "neighbor": "WM-MEM-016 Answer Grounding", "distinction": "An answer support relationship references a passage revision. Text match, provenance, retrieval rank and citation presence do not prove that an answer claim follows from the passage.", "source_refs": [ "SRC-002", "SRC-008" ] }, { "neighbor": "WM-MEM-017 Context Policy", "distinction": "Store policy identifiers and authorized decision references, not policy execution. Passage content is untrusted data and cannot override trusted context instructions.", "source_refs": [ "SRC-005", "SRC-006" ] }, { "neighbor": "Retrieval occurrence", "distinction": "Request-specific rank, score, filters, time and query references belong to an occurrence, not to the persistent passage identity.", "source_refs": [ "SRC-003", "SRC-007" ] } ] }, "sources": [ { "id": "SRC-001", "title": "Web Annotation Data Model", "organization": "W3C", "url": "https://www.w3.org/TR/annotation-model/", "version_or_date": "Recommendation, 2017-02-23", "source_type": "standard", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-10-21T17:42:27Z", "relevance": "SpecificResource, source, selector and state distinguish a selected segment from its containing representation. Sections 4.2.4-4.2.6 inform text and byte addressing; they do not prescribe retrieval chunking." }, { "id": "SRC-002", "title": "PROV-O: The PROV Ontology", "organization": "W3C", "url": "https://www.w3.org/TR/prov-o/", "version_or_date": "Recommendation, 2013-04-30", "source_type": "ontology", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-10-21T17:42:27Z", "relevance": "Entity, activity, derivation, generation, revision and invalidation support lineage. Local passage identity and lifecycle are proposed applications, not normative PROV retrieval semantics." }, { "id": "SRC-003", "title": "searchRetrieve Part 3: SRU 2.0 Version 1.0", "organization": "OASIS", "url": "https://docs.oasis-open.org/search-ws/searchRetrieve/v1.0/os/part3-sru2.0/searchRetrieve-v1.0-os-part3-sru2.0.html", "version_or_date": "OASIS Standard, 2013-01-30", "source_type": "standard", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-10-21T17:42:27Z", "relevance": "Response recordIdentifier and recordPosition distinguish retrievable record identity from result-set position. A conceptual alignment only; no claim that a passage is an SRU record without a binding." }, { "id": "SRC-004", "title": "RFC 3339: Date and Time on the Internet: Timestamps", "organization": "IETF", "url": "https://www.rfc-editor.org/rfc/rfc3339", "version_or_date": "July 2002; update RFC 9557 noted, not profiled", "source_type": "standard", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-10-21T17:42:27Z", "relevance": "Section 5.6 supplies timestamp syntax with seconds and a numeric offset or Z. Event and observation separation is a local provenance rule." }, { "id": "SRC-005", "title": "ODRL Information Model 2.2", "organization": "W3C", "url": "https://www.w3.org/TR/odrl-model/", "version_or_date": "Recommendation, 2018-02-15", "source_type": "standard", "primary_source": true, "authority_tier": 1, "accessed_at": "2026-10-21T17:42:27Z", "relevance": "Permissions, prohibitions, duties and constraints express usage policies. Expression is distinct from enforcement, rights ownership and legal applicability." }, { "id": "SRC-006", "title": "LLM01:2025 Prompt Injection", "organization": "OWASP", "url": "https://genai.owasp.org/llmrisk/llm01-prompt-injection/", "version_or_date": "2025 edition; mutable page accessed 2026-10-21", "source_type": "first-party-doc", "primary_source": true, "authority_tier": 2, "accessed_at": "2026-10-21T17:42:27Z", "relevance": "External content can carry indirect instructions. Supports policy-level separation of retrieved content and trusted instructions; it does not establish a complete defense or certify passage safety." }, { "id": "SRC-007", "title": "Apache Lucene KnnFloatVectorField API", "organization": "Apache Lucene", "url": "https://lucene.apache.org/core/10_3_1/core/org/apache/lucene/document/KnnFloatVectorField.html", "version_or_date": "10.3.1; experimental API", "source_type": "first-party-doc", "primary_source": true, "authority_tier": 2, "accessed_at": "2026-10-21T17:42:27Z", "relevance": "A vector field has dimension and similarity semantics, distinct from its document. This implementation example motivates optional representation bindings, not a required storage engine or universal vector limit." }, { "id": "SRC-008", "title": "Data on the Web Best Practices: Data Quality Vocabulary", "organization": "W3C", "url": "https://www.w3.org/TR/vocab-dqv/", "version_or_date": "Working Group Note, 2016-12-15", "source_type": "ontology", "primary_source": true, "authority_tier": 2, "accessed_at": "2026-10-21T17:42:27Z", "relevance": "Quality measurements identify a resource, metric and value. This non-normative vocabulary informs qualified passage assessments, not universal accuracy thresholds or truth certification." } ], "structure": { "bundles": [ { "id": "bundle-identity", "name": "Identity and class", "description": "Passage-specific identity and class concerns.", "rationale": "Groups passage records with shared identity and governance decisions while leaving external operations in their masters.", "source_refs": [ "SRC-002", "SRC-003", "SRC-004" ], "layers": [ { "id": "layer-identity", "name": "Persistent identity", "description": "Keep passage identity separate from a source URL, digest, index row address or result position. Equal text in distinct source contexts need not be the same passage.", "source_refs": [ "SRC-002", "SRC-003" ], "findings": [ { "id": "passage-identity", "name": "Persistent identity", "description": "Keep passage identity separate from a source URL, digest, index row address or result position. Equal text in distinct source contexts need not be the same passage. This is a proposed local modeling rule informed by the cited sources.", "source_refs": [ "SRC-002", "SRC-003" ], "questions": [ { "id": "passage-identity-q1", "kind": "identity", "text": "Which master identifier and namespace identify this passage independently of an index?", "answer_data": [ "passage_id", "namespace", "master_reference" ] }, { "id": "passage-identity-q2", "kind": "classification", "text": "Is this unit an extract, transformed text or an explicitly declared whole-source passage?", "answer_data": [ "passage_class", "extent_kind" ] }, { "id": "passage-identity-q3", "kind": "relationship", "text": "What evidence justifies equivalence or distinctness when another passage has identical text?", "answer_data": [ "related_passage_refs", "relation_kind", "identity_decision" ] } ], "data_elements": [ { "id": "passage-identity-data-passage-id", "name": "Passage identifier", "description": "Candidate passage identifier for the persistent identity finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "identifier", "cardinality": "1", "required": true, "source_refs": [ "SRC-002", "SRC-003" ] }, { "id": "passage-identity-data-class", "name": "Passage class", "description": "Candidate passage class for the persistent identity finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "code", "cardinality": "1", "required": true, "source_refs": [ "SRC-002", "SRC-003" ] }, { "id": "passage-identity-data-identity-links", "name": "Identity decision references", "description": "Candidate identity decision references for the persistent identity finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "collection", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-002", "SRC-003" ] } ], "artifacts": [ { "id": "passage-identity-artifact", "name": "Identity record", "description": "Versioned evidence for persistent identity, containing this finding's candidate answers and references, subject to access and retention rules.", "media_or_form": [ "application/json", "reference record" ], "serial": true, "identity_strategy": "Authoritative master-system record identifier, otherwise governed IRI or assigned UUID; revision and passage reference separate from time and digest.", "source_refs": [ "SRC-002", "SRC-003" ] } ], "inline_only_rationale": null } ] }, { "id": "layer-revision", "name": "Revision continuity", "description": "Use an immutable content revision and a separately versioned metadata record. A changed source state or boundary requires a successor even when the visible text happens to match.", "source_refs": [ "SRC-002", "SRC-004" ], "findings": [ { "id": "passage-revision", "name": "Revision continuity", "description": "Use an immutable content revision and a separately versioned metadata record. A changed source state or boundary requires a successor even when the visible text happens to match. This is a proposed local modeling rule informed by the cited sources.", "source_refs": [ "SRC-002", "SRC-004" ], "questions": [ { "id": "passage-revision-q1", "kind": "lifecycle", "text": "Which passage revision supersedes this one and what change caused the transition?", "answer_data": [ "revision_id", "predecessor_ref", "change_reason" ] }, { "id": "passage-revision-q2", "kind": "temporal", "text": "When was this revision generated and when was its existence observed?", "answer_data": [ "generated_at", "observed_at", "time_precision" ] }, { "id": "passage-revision-q3", "kind": "constraint", "text": "Which metadata corrections can preserve content identity under the adoption profile?", "answer_data": [ "correction_rule_ref", "expected_metadata_revision", "correction_record" ] } ], "data_elements": [ { "id": "passage-revision-data-revision-id", "name": "Content revision identifier", "description": "Candidate content revision identifier for the revision continuity finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "identifier", "cardinality": "1", "required": true, "source_refs": [ "SRC-002", "SRC-004" ] }, { "id": "passage-revision-data-predecessor", "name": "Predecessor revision", "description": "Candidate predecessor revision for the revision continuity finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "reference", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-002", "SRC-004" ] }, { "id": "passage-revision-data-times", "name": "Generation and observation times", "description": "Candidate generation and observation times for the revision continuity finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "object", "cardinality": "1", "required": true, "source_refs": [ "SRC-002", "SRC-004" ] } ], "artifacts": [ { "id": "passage-revision-artifact", "name": "Revision lineage record", "description": "Versioned evidence for revision continuity, containing this finding's candidate answers and references, subject to access and retention rules.", "media_or_form": [ "application/json", "reference record" ], "serial": true, "identity_strategy": "Authoritative master-system record identifier, otherwise governed IRI or assigned UUID; revision and passage reference separate from time and digest.", "source_refs": [ "SRC-002", "SRC-004" ] } ], "inline_only_rationale": null } ] } ] }, { "id": "bundle-anchoring", "name": "Source anchoring", "description": "Passage-specific source anchoring concerns.", "rationale": "Groups passage records with shared identity and governance decisions while leaving external operations in their masters.", "source_refs": [ "SRC-001", "SRC-002" ], "layers": [ { "id": "layer-source", "name": "Source representation", "description": "Resolve lineage to one source representation, not merely the latest URL response. References may survive loss of access, but cannot be described as currently verified without evidence.", "source_refs": [ "SRC-001", "SRC-002" ], "findings": [ { "id": "passage-source", "name": "Source representation", "description": "Resolve lineage to one source representation, not merely the latest URL response. References may survive loss of access, but cannot be described as currently verified without evidence. This is a proposed local modeling rule informed by the cited sources.", "source_refs": [ "SRC-001", "SRC-002" ], "questions": [ { "id": "passage-source-q1", "kind": "provenance", "text": "Which source master and representation revision generated the passage?", "answer_data": [ "source_master_ref", "representation_ref", "source_revision" ] }, { "id": "passage-source-q2", "kind": "evidence", "text": "What retained evidence identifies the exact source state when its URL changes?", "answer_data": [ "snapshot_ref", "source_digest", "digest_algorithm", "source_state" ] }, { "id": "passage-source-q3", "kind": "exception", "text": "How is unavailable or unverifiable source evidence recorded before retrieval eligibility is decided?", "answer_data": [ "resolution_status", "failure_reason", "last_checked_at", "eligibility_decision_ref" ] } ], "data_elements": [ { "id": "passage-source-data-source-ref", "name": "Source representation reference", "description": "Candidate source representation reference for the source representation finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "reference", "cardinality": "1", "required": true, "source_refs": [ "SRC-001", "SRC-002" ] }, { "id": "passage-source-data-source-state", "name": "Source revision evidence", "description": "Candidate source revision evidence for the source representation finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "object", "cardinality": "1", "required": true, "source_refs": [ "SRC-001", "SRC-002" ] }, { "id": "passage-source-data-resolution", "name": "Source resolution status", "description": "Candidate source resolution status for the source representation finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "code", "cardinality": "1", "required": true, "source_refs": [ "SRC-001", "SRC-002" ] } ], "artifacts": [ { "id": "passage-source-artifact", "name": "Source binding record", "description": "Versioned evidence for source representation, containing this finding's candidate answers and references, subject to access and retention rules.", "media_or_form": [ "application/json", "reference record" ], "serial": true, "identity_strategy": "Authoritative master-system record identifier, otherwise governed IRI or assigned UUID; revision and passage reference separate from time and digest.", "source_refs": [ "SRC-001", "SRC-002" ] } ], "inline_only_rationale": null } ] }, { "id": "layer-selector", "name": "Location and extent", "description": "Declare the coordinate system and extraction normalization. Text position uses a start-inclusive, end-exclusive range over Unicode code points; byte ranges use a different selector. A whole-source extent must be explicit.", "source_refs": [ "SRC-001" ], "findings": [ { "id": "passage-selector", "name": "Location and extent", "description": "Declare the coordinate system and extraction normalization. Text position uses a start-inclusive, end-exclusive range over Unicode code points; byte ranges use a different selector. A whole-source extent must be explicit. This is a proposed local modeling rule informed by the cited sources.", "source_refs": [ "SRC-001" ], "questions": [ { "id": "passage-selector-q1", "kind": "spatial", "text": "Which selector or whole-source declaration identifies the passage extent in the pinned representation?", "answer_data": [ "selector_type", "selector_value", "coordinate_space", "extent_kind" ] }, { "id": "passage-selector-q2", "kind": "validation", "text": "How are repeated quote matches, stale offsets or out-of-range boundaries detected and reported?", "answer_data": [ "match_count", "validation_outcome", "checked_source_revision" ] }, { "id": "passage-selector-q3", "kind": "interoperability", "text": "Which normalization and offset conventions must an importing system preserve?", "answer_data": [ "normalization_profile_ref", "offset_unit", "end_convention", "mapping_loss" ] } ], "data_elements": [ { "id": "passage-selector-data-selector", "name": "Selector or whole-source extent", "description": "Candidate selector or whole-source extent for the location and extent finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "object", "cardinality": "1", "required": true, "source_refs": [ "SRC-001" ] }, { "id": "passage-selector-data-normalization", "name": "Normalization profile", "description": "Candidate normalization profile for the location and extent finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "reference", "cardinality": "1", "required": true, "source_refs": [ "SRC-001" ] }, { "id": "passage-selector-data-anchor-check", "name": "Anchor resolution evidence", "description": "Candidate anchor resolution evidence for the location and extent finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "object", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-001" ] } ], "artifacts": [ { "id": "passage-selector-artifact", "name": "Anchor validation record", "description": "Versioned evidence for location and extent, containing this finding's candidate answers and references, subject to access and retention rules.", "media_or_form": [ "application/json", "reference record" ], "serial": true, "identity_strategy": "Authoritative master-system record identifier, otherwise governed IRI or assigned UUID; revision and passage reference separate from time and digest.", "source_refs": [ "SRC-001" ] } ], "inline_only_rationale": null } ] } ] }, { "id": "bundle-representation", "name": "Content and derivation", "description": "Passage-specific content and derivation concerns.", "rationale": "Groups passage records with shared identity and governance decisions while leaving external operations in their masters.", "source_refs": [ "SRC-001", "SRC-002", "SRC-005", "SRC-008" ], "layers": [ { "id": "layer-payload", "name": "Payload properties", "description": "A passage can retain text or only an authorized locator. Media type, language, encoding and extent describe its representation; token counts require a tokenizer binding and are not universal sizes.", "source_refs": [ "SRC-001", "SRC-005" ], "findings": [ { "id": "passage-payload", "name": "Payload properties", "description": "A passage can retain text or only an authorized locator. Media type, language, encoding and extent describe its representation; token counts require a tokenizer binding and are not universal sizes. This is a proposed local modeling rule informed by the cited sources.", "source_refs": [ "SRC-001", "SRC-005" ], "questions": [ { "id": "passage-payload-q1", "kind": "definition", "text": "What text or authorized content locator constitutes this passage revision?", "answer_data": [ "payload_mode", "payload_or_locator", "media_type", "encoding" ] }, { "id": "passage-payload-q2", "kind": "measurement", "text": "Which character, byte or token counts were measured with which method?", "answer_data": [ "count_value", "count_unit", "tokenizer_ref", "measured_revision" ] }, { "id": "passage-payload-q3", "kind": "privacy", "text": "What information is removed or withheld from the stored or exported representation?", "answer_data": [ "redaction_profile_ref", "withheld_fields", "authorized_view_ref" ] } ], "data_elements": [ { "id": "passage-payload-data-payload", "name": "Text or reference payload", "description": "Candidate text or reference payload for the payload properties finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "object", "cardinality": "1", "required": true, "source_refs": [ "SRC-001", "SRC-005" ] }, { "id": "passage-payload-data-language", "name": "Language declarations", "description": "Candidate language declarations for the payload properties finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "collection", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-001", "SRC-005" ] }, { "id": "passage-payload-data-size", "name": "Representation size measurements", "description": "Candidate representation size measurements for the payload properties finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "collection", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-001", "SRC-005" ] } ], "artifacts": [ { "id": "passage-payload-artifact", "name": "Representation manifest", "description": "Versioned evidence for payload properties, containing this finding's candidate answers and references, subject to access and retention rules.", "media_or_form": [ "application/json", "reference record" ], "serial": true, "identity_strategy": "Authoritative master-system record identifier, otherwise governed IRI or assigned UUID; revision and passage reference separate from time and digest.", "source_refs": [ "SRC-001", "SRC-005" ] } ], "inline_only_rationale": null } ] }, { "id": "layer-derivation", "name": "Extraction lineage", "description": "Record the external extraction activity, its inputs and profile. OCR, translation, normalization and summarization are declared transformations, not silently described as verbatim quotation.", "source_refs": [ "SRC-002", "SRC-008" ], "findings": [ { "id": "passage-derivation", "name": "Extraction lineage", "description": "Record the external extraction activity, its inputs and profile. OCR, translation, normalization and summarization are declared transformations, not silently described as verbatim quotation. This is a proposed local modeling rule informed by the cited sources.", "source_refs": [ "SRC-002", "SRC-008" ], "questions": [ { "id": "passage-derivation-q1", "kind": "process", "text": "Which extraction or transformation activity produced the passage and under which profile?", "answer_data": [ "activity_ref", "transform_kind", "profile_revision", "responsible_role_ref" ] }, { "id": "passage-derivation-q2", "kind": "composition", "text": "Which contextual source spans were included, overlapped or omitted in the extraction?", "answer_data": [ "included_spans", "overlap_refs", "omitted_context", "source_representation_ref" ] }, { "id": "passage-derivation-q3", "kind": "quality", "text": "Which observations quantify transformation loss or unresolved extraction errors?", "answer_data": [ "quality_observation_refs", "metric_ref", "uncertainty", "review_state" ] } ], "data_elements": [ { "id": "passage-derivation-data-activity", "name": "Derivation activity reference", "description": "Candidate derivation activity reference for the extraction lineage finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "reference", "cardinality": "1", "required": true, "source_refs": [ "SRC-002", "SRC-008" ] }, { "id": "passage-derivation-data-transform", "name": "Transformation declarations", "description": "Candidate transformation declarations for the extraction lineage finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "collection", "cardinality": "1..n", "required": true, "source_refs": [ "SRC-002", "SRC-008" ] }, { "id": "passage-derivation-data-context", "name": "Context and overlap references", "description": "Candidate context and overlap references for the extraction lineage finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "collection", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-002", "SRC-008" ] } ], "artifacts": [ { "id": "passage-derivation-artifact", "name": "Derivation record", "description": "Versioned evidence for extraction lineage, containing this finding's candidate answers and references, subject to access and retention rules.", "media_or_form": [ "application/json", "reference record" ], "serial": true, "identity_strategy": "Authoritative master-system record identifier, otherwise governed IRI or assigned UUID; revision and passage reference separate from time and digest.", "source_refs": [ "SRC-002", "SRC-008" ] } ], "inline_only_rationale": null } ] } ] }, { "id": "bundle-retrievability", "name": "Retrieval bindings", "description": "Passage-specific retrieval bindings concerns.", "rationale": "Groups passage records with shared identity and governance decisions while leaving external operations in their masters.", "source_refs": [ "SRC-003", "SRC-007", "SRC-008" ], "layers": [ { "id": "layer-index", "name": "Index representation binding", "description": "Reference index-specific representations without making embeddings mandatory. Each optional vector binding identifies its model revision, dimensions and comparison convention; re-embedding does not itself replace passage content.", "source_refs": [ "SRC-003", "SRC-007" ], "findings": [ { "id": "passage-index", "name": "Index representation binding", "description": "Reference index-specific representations without making embeddings mandatory. Each optional vector binding identifies its model revision, dimensions and comparison convention; re-embedding does not itself replace passage content. This is a proposed local modeling rule informed by the cited sources.", "source_refs": [ "SRC-003", "SRC-007" ], "questions": [ { "id": "passage-index-q1", "kind": "relationship", "text": "Which index snapshots contain a representation of this exact passage revision?", "answer_data": [ "index_ref", "index_snapshot", "representation_ref", "passage_revision" ] }, { "id": "passage-index-q2", "kind": "constraint", "text": "If a vector representation exists, which model, dimensions and similarity convention define compatibility?", "answer_data": [ "embedding_model_ref", "model_revision", "dimensions", "similarity", "normalization_rule" ] }, { "id": "passage-index-q3", "kind": "state", "text": "How is a stale or mismatched index representation marked before it can be used?", "answer_data": [ "binding_status", "mismatch_reason", "checked_at", "index_owner_ref" ] } ], "data_elements": [ { "id": "passage-index-data-index-bindings", "name": "Index membership bindings", "description": "Candidate index membership bindings for the index representation binding finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "collection", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-003", "SRC-007" ] }, { "id": "passage-index-data-vector-bindings", "name": "Optional vector metadata", "description": "Candidate optional vector metadata for the index representation binding finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "collection", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-003", "SRC-007" ] }, { "id": "passage-index-data-binding-status", "name": "Representation compatibility evidence", "description": "Candidate representation compatibility evidence for the index representation binding finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "collection", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-003", "SRC-007" ] } ], "artifacts": [ { "id": "passage-index-artifact", "name": "Index binding manifest", "description": "Versioned evidence for index representation binding, containing this finding's candidate answers and references, subject to access and retention rules.", "media_or_form": [ "application/json", "reference record" ], "serial": true, "identity_strategy": "Authoritative master-system record identifier, otherwise governed IRI or assigned UUID; revision and passage reference separate from time and digest.", "source_refs": [ "SRC-003", "SRC-007" ] } ], "inline_only_rationale": null } ] }, { "id": "layer-occurrence", "name": "Retrieval occurrence references", "description": "Link to retrieval events that returned this revision. Rank and score remain qualified by request, index snapshot and scoring method; they are not stable properties or truth probabilities.", "source_refs": [ "SRC-003", "SRC-007", "SRC-008" ], "findings": [ { "id": "passage-occurrence", "name": "Retrieval occurrence references", "description": "Link to retrieval events that returned this revision. Rank and score remain qualified by request, index snapshot and scoring method; they are not stable properties or truth probabilities. This is a proposed local modeling rule informed by the cited sources.", "source_refs": [ "SRC-003", "SRC-007", "SRC-008" ], "questions": [ { "id": "passage-occurrence-q1", "kind": "event", "text": "Which retrieval occurrence returned this passage revision and through which index snapshot?", "answer_data": [ "occurrence_ref", "request_ref", "index_snapshot", "returned_revision" ] }, { "id": "passage-occurrence-q2", "kind": "measurement", "text": "What score and rank were reported under which scoring scale or method?", "answer_data": [ "score", "rank", "score_method_ref", "scale", "occurrence_ref" ] }, { "id": "passage-occurrence-q3", "kind": "evidence", "text": "Which downstream citation references this exact revision without asserting claim entailment here?", "answer_data": [ "grounding_ref", "cited_revision", "citation_locator", "support_review_ref" ] } ], "data_elements": [ { "id": "passage-occurrence-data-occurrences", "name": "Retrieval occurrence references", "description": "Candidate retrieval occurrence references for the retrieval occurrence references finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "collection", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-003", "SRC-007", "SRC-008" ] }, { "id": "passage-occurrence-data-citations", "name": "Downstream citation references", "description": "Candidate downstream citation references for the retrieval occurrence references finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "collection", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-003", "SRC-007", "SRC-008" ] } ], "artifacts": [ { "id": "passage-occurrence-artifact", "name": "Occurrence linkage record", "description": "Versioned evidence for retrieval occurrence references, containing this finding's candidate answers and references, subject to access and retention rules.", "media_or_form": [ "application/json", "reference record" ], "serial": true, "identity_strategy": "Authoritative master-system record identifier, otherwise governed IRI or assigned UUID; revision and passage reference separate from time and digest.", "source_refs": [ "SRC-003", "SRC-007", "SRC-008" ] } ], "inline_only_rationale": null } ] } ] }, { "id": "bundle-assurance", "name": "Quality and trust", "description": "Passage-specific quality and trust concerns.", "rationale": "Groups passage records with shared identity and governance decisions while leaving external operations in their masters.", "source_refs": [ "SRC-002", "SRC-005", "SRC-006", "SRC-008" ], "layers": [ { "id": "layer-quality", "name": "Fitness observations", "description": "Record checks of anchoring, extraction and use fitness as observations with defined metrics and evidence. Unknown, failed and not evaluated remain distinct; quality assessments can disagree.", "source_refs": [ "SRC-008", "SRC-002" ], "findings": [ { "id": "passage-quality", "name": "Fitness observations", "description": "Record checks of anchoring, extraction and use fitness as observations with defined metrics and evidence. Unknown, failed and not evaluated remain distinct; quality assessments can disagree. This is a proposed local modeling rule informed by the cited sources.", "source_refs": [ "SRC-008", "SRC-002" ], "questions": [ { "id": "passage-quality-q1", "kind": "quality", "text": "Which fitness criterion was measured for this revision and by what method?", "answer_data": [ "metric_ref", "method_ref", "value", "unit", "evaluated_revision" ] }, { "id": "passage-quality-q2", "kind": "provenance", "text": "Who made the assessment and what evidence and observation time support it?", "answer_data": [ "assessor_role_ref", "evidence_ref", "observed_at" ] }, { "id": "passage-quality-q3", "kind": "decision", "text": "How are contradictory assessments or insufficient evidence represented without inventing an approval?", "answer_data": [ "assessment_refs", "disagreement_reason", "resolution_state", "decision_ref" ] } ], "data_elements": [ { "id": "passage-quality-data-assessments", "name": "Quality observations", "description": "Candidate quality observations for the fitness observations finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "collection", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-008", "SRC-002" ] }, { "id": "passage-quality-data-assessment-state", "name": "Quality review state", "description": "Candidate quality review state for the fitness observations finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "code", "cardinality": "1", "required": true, "source_refs": [ "SRC-008", "SRC-002" ] } ], "artifacts": [ { "id": "passage-quality-artifact", "name": "Quality observation report", "description": "Versioned evidence for fitness observations, containing this finding's candidate answers and references, subject to access and retention rules.", "media_or_form": [ "application/json", "reference record" ], "serial": true, "identity_strategy": "Authoritative master-system record identifier, otherwise governed IRI or assigned UUID; revision and passage reference separate from time and digest.", "source_refs": [ "SRC-008", "SRC-002" ] } ], "inline_only_rationale": null } ] }, { "id": "layer-trust", "name": "Untrusted content handling", "description": "Treat passage text and attached metadata as untrusted input. Store trust labels and review decisions at policy level; neither a digest nor a screening result grants instruction authority or guarantees safety.", "source_refs": [ "SRC-006", "SRC-005" ], "findings": [ { "id": "passage-trust", "name": "Untrusted content handling", "description": "Treat passage text and attached metadata as untrusted input. Store trust labels and review decisions at policy level; neither a digest nor a screening result grants instruction authority or guarantees safety. This is a proposed local modeling rule informed by the cited sources.", "source_refs": [ "SRC-006", "SRC-005" ], "questions": [ { "id": "passage-trust-q1", "kind": "security", "text": "Which trust boundary prevents instructions in retrieved content from becoming authorized commands?", "answer_data": [ "trust_class", "boundary_policy_ref", "permitted_use" ] }, { "id": "passage-trust-q2", "kind": "authority", "text": "Which role can release quarantined content under the applicable review policy?", "answer_data": [ "review_role_ref", "release_decision_ref", "scope", "expiry" ] }, { "id": "passage-trust-q3", "kind": "exception", "text": "What policy response applies when suspicious content or unsupported screening claims are encountered?", "answer_data": [ "quarantine_state", "reason_code", "escalation_ref", "screening_evidence_ref" ] } ], "data_elements": [ { "id": "passage-trust-data-trust", "name": "Trust classification", "description": "Candidate trust classification for the untrusted content handling finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "code", "cardinality": "1", "required": true, "source_refs": [ "SRC-006", "SRC-005" ] }, { "id": "passage-trust-data-security-policy", "name": "Content handling policy reference", "description": "Candidate content handling policy reference for the untrusted content handling finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "reference", "cardinality": "1", "required": true, "source_refs": [ "SRC-006", "SRC-005" ] }, { "id": "passage-trust-data-screening", "name": "Screening and release references", "description": "Candidate screening and release references for the untrusted content handling finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "collection", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-006", "SRC-005" ] } ], "artifacts": [ { "id": "passage-trust-artifact", "name": "Trust disposition record", "description": "Versioned evidence for untrusted content handling, containing this finding's candidate answers and references, subject to access and retention rules.", "media_or_form": [ "application/json", "reference record" ], "serial": true, "identity_strategy": "Authoritative master-system record identifier, otherwise governed IRI or assigned UUID; revision and passage reference separate from time and digest.", "source_refs": [ "SRC-006", "SRC-005" ] } ], "inline_only_rationale": null } ] } ] }, { "id": "bundle-governance", "name": "Use and continuity", "description": "Passage-specific use and continuity concerns.", "rationale": "Groups passage records with shared identity and governance decisions while leaving external operations in their masters.", "source_refs": [ "SRC-002", "SRC-004", "SRC-005", "SRC-006" ], "layers": [ { "id": "layer-rights", "name": "Rights and access binding", "description": "Bind source rights, passage permissions and context policy without granting rights from a citation. The serving system must evaluate current authority before returning text, locators or revealing metadata.", "source_refs": [ "SRC-005", "SRC-006" ], "findings": [ { "id": "passage-rights", "name": "Rights and access binding", "description": "Bind source rights, passage permissions and context policy without granting rights from a citation. The serving system must evaluate current authority before returning text, locators or revealing metadata. This is a proposed local modeling rule informed by the cited sources.", "source_refs": [ "SRC-005", "SRC-006" ], "questions": [ { "id": "passage-rights-q1", "kind": "ownership", "text": "Which accountable roles maintain this passage and who supplies the rights basis?", "answer_data": [ "maintainer_role_ref", "rights_authority_ref", "license_or_policy_ref" ] }, { "id": "passage-rights-q2", "kind": "access", "text": "Which audience, purpose and policy decision authorize the requested passage view?", "answer_data": [ "audience_scope", "purpose", "decision_ref", "decision_expiry", "view_ref" ] }, { "id": "passage-rights-q3", "kind": "requirement", "text": "What attribution, usage restrictions and duties accompany an authorized export?", "answer_data": [ "permission_refs", "prohibition_refs", "duty_refs", "attribution_fields" ] } ], "data_elements": [ { "id": "passage-rights-data-owner-role", "name": "Accountable maintainer role", "description": "Candidate accountable maintainer role for the rights and access binding finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "reference", "cardinality": "1", "required": true, "source_refs": [ "SRC-005", "SRC-006" ] }, { "id": "passage-rights-data-policy-bindings", "name": "Rights and access policies", "description": "Candidate rights and access policies for the rights and access binding finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "collection", "cardinality": "1..n", "required": true, "source_refs": [ "SRC-005", "SRC-006" ] }, { "id": "passage-rights-data-use-decision", "name": "Authorized use decision reference", "description": "Candidate authorized use decision reference for the rights and access binding finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "reference", "cardinality": "0..1", "required": false, "source_refs": [ "SRC-005", "SRC-006" ] } ], "artifacts": [ { "id": "passage-rights-artifact", "name": "Use policy binding record", "description": "Versioned evidence for rights and access binding, containing this finding's candidate answers and references, subject to access and retention rules.", "media_or_form": [ "application/json", "reference record" ], "serial": true, "identity_strategy": "Authoritative master-system record identifier, otherwise governed IRI or assigned UUID; revision and passage reference separate from time and digest.", "source_refs": [ "SRC-005", "SRC-006" ] } ], "inline_only_rationale": null } ] }, { "id": "layer-disposition", "name": "Availability and disposition", "description": "Distinguish active, unverified, quarantined, superseded, withdrawn and disposed states under an adoption profile. Local withdrawal is not proof that every external index, cache or answer has been purged.", "source_refs": [ "SRC-002", "SRC-004", "SRC-005" ], "findings": [ { "id": "passage-disposition", "name": "Availability and disposition", "description": "Distinguish active, unverified, quarantined, superseded, withdrawn and disposed states under an adoption profile. Local withdrawal is not proof that every external index, cache or answer has been purged. This is a proposed local modeling rule informed by the cited sources.", "source_refs": [ "SRC-002", "SRC-004", "SRC-005" ], "questions": [ { "id": "passage-disposition-q1", "kind": "state", "text": "What availability state and evidence determine whether this revision may be served?", "answer_data": [ "availability_state", "reason", "evidence_ref", "effective_at" ] }, { "id": "passage-disposition-q2", "kind": "retention", "text": "Which retention rule or authorized hold controls payload disposal and minimal tombstone metadata?", "answer_data": [ "retention_policy_ref", "hold_ref", "disposition_due", "tombstone_fields" ] }, { "id": "passage-disposition-q3", "kind": "lifecycle", "text": "Which external index or cache owners acknowledged withdrawal and which remain pending?", "answer_data": [ "withdrawal_request_refs", "acknowledgment_refs", "pending_targets", "completion_scope" ] } ], "data_elements": [ { "id": "passage-disposition-data-availability", "name": "Passage availability state", "description": "Candidate passage availability state for the availability and disposition finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "code", "cardinality": "1", "required": true, "source_refs": [ "SRC-002", "SRC-004", "SRC-005" ] }, { "id": "passage-disposition-data-retention", "name": "Disposition policy binding", "description": "Candidate disposition policy binding for the availability and disposition finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "reference", "cardinality": "1", "required": true, "source_refs": [ "SRC-002", "SRC-004", "SRC-005" ] }, { "id": "passage-disposition-data-withdrawal", "name": "External withdrawal acknowledgments", "description": "Candidate external withdrawal acknowledgments for the availability and disposition finding; profile must define nested fields, allowed values and null semantics.", "value_kind": "collection", "cardinality": "0..n", "required": false, "source_refs": [ "SRC-002", "SRC-004", "SRC-005" ] } ], "artifacts": [ { "id": "passage-disposition-artifact", "name": "Disposition record", "description": "Versioned evidence for availability and disposition, containing this finding's candidate answers and references, subject to access and retention rules.", "media_or_form": [ "application/json", "reference record" ], "serial": true, "identity_strategy": "Authoritative master-system record identifier, otherwise governed IRI or assigned UUID; revision and passage reference separate from time and digest.", "source_refs": [ "SRC-002", "SRC-004", "SRC-005" ] } ], "inline_only_rationale": null } ] } ] } ] }, "functions": [ { "id": "register-passage", "name": "Register passage revision", "description": "Proposed local operation; not implemented. Record identity and lineage for supplied authorized content without running an extractor.", "inputs": [ "Source revision and extent evidence", "Passage identity and payload or locator", "Derivation and policy references" ], "outputs": [ "Draft passage revision or rejection reasons" ], "preconditions": [ "Authorized maintainer", "Identity and source reference present", "Expected revision agrees" ], "effects": [ "Creates a local unverified record; no indexing or serving authorization" ], "source_refs": [ "SRC-001", "SRC-002", "SRC-005" ] }, { "id": "check-anchor", "name": "Record anchor check", "description": "Proposed local operation; not implemented. Compare supplied permitted source representation and selectors under the declared profile.", "inputs": [ "Pinned source representation", "Selector, normalization and expected passage revision" ], "outputs": [ "Unique match, ambiguous match, mismatch or unavailable result" ], "preconditions": [ "Read authority for both resources", "Declared coordinate and normalization profile" ], "effects": [ "Appends local evidence; ambiguous matches do not silently select the first occurrence" ], "source_refs": [ "SRC-001", "SRC-008" ] }, { "id": "bind-index", "name": "Record index representation", "description": "Proposed local operation; not implemented. Bind an existing representation reported by the external index master.", "inputs": [ "Index snapshot and representation reference", "Passage revision", "Optional vector compatibility metadata" ], "outputs": [ "Binding record or incompatible result" ], "preconditions": [ "Authorized binding role", "Exact passage revision match", "Index report available" ], "effects": [ "Records local reference only; does not embed, build, search or alter an index" ], "source_refs": [ "SRC-003", "SRC-007" ] }, { "id": "attach-assessment", "name": "Attach quality observation", "description": "Proposed local operation; not implemented. Associate supplied quality and trust review evidence with an exact revision.", "inputs": [ "Assessment reference and metric", "Evaluated revision", "Assessor and observation time" ], "outputs": [ "Qualified assessment link or refusal" ], "preconditions": [ "Assessment evidence available to authorized reviewer", "Metric and revision resolved" ], "effects": [ "Preserves contradictory and unknown outcomes; does not certify truth or safety" ], "source_refs": [ "SRC-002", "SRC-006", "SRC-008" ] }, { "id": "prepare-view", "name": "Prepare authorized passage view", "description": "Proposed local operation; not implemented. Project allowed content and citation metadata using a supplied current authorization decision.", "inputs": [ "Audience and purpose", "Current policy decision reference", "Passage revision and requested fields" ], "outputs": [ "Scoped local view with citation locator or refusal" ], "preconditions": [ "Decision covers requested fields and remains current", "Passage eligible and source binding verified", "No unresolved source or policy conflict" ], "effects": [ "Produces local projection with use restrictions; no external transmission or answer entailment claim" ], "source_refs": [ "SRC-001", "SRC-005", "SRC-006" ] }, { "id": "record-withdrawal", "name": "Record passage withdrawal", "description": "Proposed local operation; not implemented. Mark a local revision unavailable and record externally supplied disposition acknowledgments.", "inputs": [ "Authorized withdrawal decision", "Revision and expected state", "Retention and hold references", "External acknowledgment references" ], "outputs": [ "Local withdrawal record and outstanding target list" ], "preconditions": [ "Disposition authority verified", "Retention obligations resolved for the requested local change" ], "effects": [ "Prevents local eligibility under the profile; external deletion remains owned by index, cache and storage custodians" ], "source_refs": [ "SRC-002", "SRC-004", "SRC-005" ] } ], "composition": [ { "target": "Source representation master", "relation": "REFERENCE", "purpose": "Mandatory source identity and revision evidence; document master owns source editing and retention.", "required": true, "source_refs": [ "SRC-001", "SRC-002" ] }, { "target": "WM-MEM-014", "relation": "REFERENCE", "purpose": "Optional retrieval index and snapshot bindings; reconcile the registry parent clue without inheriting index execution.", "required": false, "source_refs": [ "SRC-003", "SRC-007" ] }, { "target": "WM-MEM-016", "relation": "REFERENCE", "purpose": "Optional answer grounding references; that model owns statement-to-evidence support judgments.", "required": false, "source_refs": [ "SRC-002", "SRC-008" ] }, { "target": "WM-MEM-017", "relation": "REFERENCE", "purpose": "Optional context policy binding when adopted; local rights policy remains mandatory even without this model binding.", "required": false, "source_refs": [ "SRC-005", "SRC-006" ] }, { "target": "W3C Web Annotation Data Model", "relation": "ALIGN", "purpose": "Conceptual alignment for source, state and selectors; a tested serialization mapping is deferred.", "required": false, "source_refs": [ "SRC-001" ] }, { "target": "W3C PROV-O", "relation": "ALIGN", "purpose": "Conceptual derivation and revision mapping; passage-specific constraints are locally proposed.", "required": false, "source_refs": [ "SRC-002" ] }, { "target": "ODRL 2.2", "relation": "ALIGN", "purpose": "Represent policy statements without claiming runtime enforcement or legal compliance.", "required": false, "source_refs": [ "SRC-005" ] } ], "serviceLayers": { "dimension": { "owner_package_requirements": [ "Identify an accountable maintainer role, namespace and passage master; no company is assigned ownership by this model.", "Pin source, index and policy bindings and declare permitted content, metadata and lineage scopes.", "Declare representation, selector, revision, retention and acceptance profiles before operational adoption." ], "namespace_guidance": "Use a governed passage namespace with stable local IDs; keep metadata revisions, content revisions, source identities and external index row IDs distinct.", "registry_links": [ "WM-MEM-015", "WM-MEM-014", "WM-MEM-016", "WM-MEM-017" ] }, "canon_and_patch": { "canonicalization_rules": [ "Research remains a noncanonical reviewable draft under a single-provider waiver.", "Pin normalization before computing a digest; identify algorithm and exact byte representation. Digest equality is not semantic identity or rights evidence." ], "patch_rules": [ "Require expected metadata revision and authorized role; reject stale writes.", "Do not overwrite content, source state or selector on an existing revision; record successor and lineage. Correct metadata with a governed history subject to retention." ], "compatibility_rules": [ "Changes to coordinate units, normalization or identity meaning require a new profile version and explicit migration mapping.", "Do not equate token offsets, Unicode code points and bytes. Lossy or ambiguous imports remain unverified." ] }, "artifact_rules": { "identity_priority": [ "Authoritative master-system identifier", "Governed global identifier or IRI", "UUID or ULID assigned by the adopting Dimension" ], "timestamp_rule": "Use RFC 3339 with seconds and an explicit offset or Z. Separate source event time, extraction time and observation or ingestion time; unknown time is not fabricated.", "serial_naming_rule": "Use stable record identifier plus revision and sequence; dates and hashes are metadata, not primary identity.", "integrity_rule": "Record digest algorithm, payload representation and verification result; protect quoted source text, locators and hashes under the same applicable policy. Integrity does not prove truth." }, "policies": [ "Treat retrieved content as untrusted data; it cannot amend policies or authorize actions. Dangerous content is handled only through policy references and disposition states.", "Evaluate current access at serving and export boundaries. Source visibility, index membership and past permission are not present authority.", "Minimize content and sensitive query references. Copy permissions, restrictions and attribution obligations with an authorized view.", "Retention and lawful disposal constrain history. Preserve only permitted tombstone metadata; referenced owners execute their own removals." ], "crud": { "read": [ "Read the authorized view only after current access and availability checks; prevent locators, scores or metadata from disclosing denied content." ], "create": [ "Allocate identity, pin source and extent, record derivation and rights; start unverified until required checks complete." ], "update": [ "Use optimistic concurrency and qualified evidence; content changes create successor revisions, index-binding changes do not replace passage identity." ], "delete": [ "Withdraw local availability first when authorized. Dispose local payload and unnecessary history only under retention and hold policy; preserve a minimal permitted tombstone.", "Index, cache, backup and source custodians own their deletion execution. Track acknowledgments and pending targets without claiming global erasure." ] }, "roles": [ { "name": "Passage maintainer", "responsibilities": [ "Own local identity, revisions and profile selection." ] }, { "name": "Source custodian", "responsibilities": [ "Supply authoritative source state and applicable rights evidence." ] }, { "name": "Index custodian", "responsibilities": [ "Own index representations, refresh and removal acknowledgments." ] }, { "name": "Evidence reviewer", "responsibilities": [ "Assess anchoring and quality evidence without asserting unsupported truth." ] }, { "name": "Policy custodian", "responsibilities": [ "Authorize use scopes, exceptions and disposition decisions." ] }, { "name": "Consumer agent", "responsibilities": [ "Read allowed passage data without treating it as instructions." ] } ], "access": { "default_rule": "Deny access unless a current scoped authorization and eligible passage state are established; metadata access is separately governed.", "scopes": [ "bundle", "layer", "finding", "artifact" ], "exceptions": [ "Time-bounded exceptions require an accountable authority, purpose and recorded scope; no exception is inferred from a source URL." ], "audit_requirements": [ "Reference access, revision, export and disposition events in the adopting audit service; retain minimal identifiers and decision references without copying sensitive payload by default." ] }, "agents_bootstrap": { "filename": "AGENTS.md", "required_fields": [ "Name", "Type", "Specification URL", "Storage type URL", "Interface URL", "Processes URL" ], "read_order": [ "AGENTS.md", "spec.yaml and research holds", "Adopting Dimension policy and pinned profiles", "Source, index and policy master bindings" ] } }, "coverage": { "claim": "Source-grounded proposed structure for one retrieval passage entity with pinned source lineage, qualified representation bindings and governed local availability. Separate frozen no-tools Codex self-audit completed. Independent review, direct source/version verification and executable adoption conformance remain holds. This is a noncanonical reviewable draft without a universal completeness claim.", "confidence": "medium", "checklist": [ { "dimension": "identity", "status": "covered", "notes": "Master passage identity differs from URL, digest, vector row and rank." }, { "dimension": "lifecycle", "status": "covered", "notes": "Content revisions, metadata corrections and withdrawal have distinct records." }, { "dimension": "relationships", "status": "covered", "notes": "Source, index, grounding and policy masters are referenced." }, { "dimension": "temporal", "status": "covered", "notes": "Event, generation and observation time remain separate." }, { "dimension": "provenance", "status": "covered", "notes": "Source state and extraction activity preserve transformation lineage." }, { "dimension": "ownership", "status": "covered", "notes": "Accountable roles separate passage, source, index and policy authority." }, { "dimension": "validation", "status": "covered", "notes": "Ambiguous selectors, mismatched vectors and uncertain evidence have explicit outcomes; implementation deferred." }, { "dimension": "access", "status": "covered", "notes": "Current policy decisions govern content and metadata views." }, { "dimension": "retention and deletion", "status": "covered", "notes": "Local withdrawal and disposal are distinct from external removal acknowledgments." }, { "dimension": "interoperability", "status": "covered", "notes": "Web Annotation, PROV-O and ODRL are conceptual alignments only." }, { "dimension": "direct properties", "status": "covered", "notes": "Payload mode, language, media type, extent and measured size are represented." }, { "dimension": "recognition and observation", "status": "covered", "notes": "Anchoring, extraction quality and trust are evidence-bearing observations." }, { "dimension": "capabilities and actions", "status": "covered", "notes": "Six proposed local operations have preconditions, refusal outputs and bounded effects." }, { "dimension": "physical properties", "status": "not-applicable", "notes": "A digital passage has no intrinsic mass or physical dimensions; storage hardware is external." }, { "dimension": "instance conformance", "status": "gap", "notes": "Nested schemas, enumeration profiles, executable mappings and adversarial fixtures are not supplied." }, { "dimension": "independent review", "status": "gap", "notes": "Local Codex self-audit is required; external providers are waived, not substituted." }, { "dimension": "source verification", "status": "gap", "notes": "Selected web-service content reviewed; direct HTTP is unattempted under the reported sandbox restriction." } ], "known_omissions": [ "Nested executable instance schemas, complete mappings and pinned neighboring model versions.", "Adoption-specific chunking rules, multilingual and bidirectional text fixtures, OCR and translation assessment benchmarks.", "Multimodal and multi-source passage profiles; external index ranking and answer entailment evaluation.", "Independent external review and direct HTTP/version/errata verification." ], "conflicts": [], "regional_assumptions": [ "No universal license, privacy, retention or export entitlement is asserted; jurisdiction and institutional profiles must be supplied by the adopter." ], "adversarial_checks": [ "Repeated quote matches cannot silently select the first occurrence.", "Identical text in different source contexts cannot silently merge identities.", "Changing the embedding model does not silently replace passage content.", "A high retrieval score does not prove truth or authorize disclosure.", "Source withdrawal does not imply removal from every external cache.", "Embedded instructions remain untrusted content even after quality screening." ] }, "researchAdjudication": { "providerMode": "single-provider-waiver", "activeProviders": [ "codex" ], "waivedProviders": [ "claude", "grok" ], "providerPolicy": { "contract_version": "1.0.0", "mode": "single-provider-waiver", "effective_at": "2026-09-06T00:00:00Z", "scope": "Canonical single-stream subject-model research after the six-workstream consolidation", "active_providers": [ "codex" ], "waived_providers": [ { "provider": "claude", "authorized_by": "repository owner", "authorized_at": "2026-09-06T00:00:00Z", "reason": "Claude produced no result on prior 1800-second and 900-second attempts and again timed out on bounded 600-second Sonnet and 300-second Haiku passes. The owner prioritized completion over provider availability." }, { "provider": "grok", "authorized_by": "repository owner", "authorized_at": "2026-09-06T00:00:00Z", "reason": "The repository owner authorized completion without Grok when Grok is unavailable, slow or schema-invalid. Grok may still be attempted as a bounded supplemental reviewer, but its failure never blocks a valid Claude plus no-tools result." } ], "review_rule": "Codex may complete source-grounded fallback research after bounded Claude and Grok attempts fail. It requires a separate no-tools adversarial audit and remains reviewable-draft with a visible absence-of-external-review hold.", "supplemental_provider_attempts": [ { "provider": "claude", "required": false, "maximum_attempts": 1, "failure_policy": "record-and-continue", "admission_rule": "Use only a locally schema-valid result whose sources and boundaries survive adjudication." }, { "provider": "grok", "required": false, "maximum_attempts": 1, "failure_policy": "record-and-continue", "admission_rule": "Use only a locally schema-valid result whose sources and boundaries survive adjudication." } ] }, "boundaryDecision": { "entry_kind": "entity", "status": "accepted", "rationale": "The passage has persistent identity distinct from its source representation, content revision, index row and query occurrence. The registry standalone-mm classifies the record plane. Its parent index clue supports a qualified reference, not inherited index execution or lifecycle ownership." }, "decisions": [ { "concept": "Entity and registry parent", "disposition": "accepted with qualification", "rationale": "Passage identity persists across metadata changes and index membership. The empty relation ledger supplies no operational inheritance; the parent clue is represented as an optional index reference." }, { "concept": "Source revision and passage identity", "disposition": "accepted", "rationale": "Source state and extent changes create successors even when text matches. Digest and URL equality do not collapse distinct source contexts or replace the master identifier." }, { "concept": "Whole-source and reference-only cases", "disposition": "accepted", "rationale": "Explicit extent and payload modes handle these cases without requiring a copied excerpt. The one-source core remains distinct from multi-source generated answers." }, { "concept": "Selector ambiguity", "disposition": "qualified local rule", "rationale": "Code point and byte addressing are separated. Refusing ambiguous serving is a local profile decision, explicitly stronger than a selector that can match several locations; it is not misrepresented as a normative requirement." }, { "concept": "Transformations and quotation", "disposition": "separated", "rationale": "Extraction lineage records OCR, translation and other transformations. Such outputs cannot silently be treated as verbatim quotation or as the authoritative source document." }, { "concept": "Embedding and index ownership", "disposition": "separated", "rationale": "Optional vector metadata and index memberships do not define the passage root. The local binding function neither generates embeddings nor owns indexing or refresh operations." }, { "concept": "Retrieval score and grounding", "disposition": "separated", "rationale": "Request-scoped rank and score remain on occurrence references. Citation and quality evidence do not assert answer entailment, truth probability or independent verification." }, { "concept": "Trust and authority", "disposition": "accepted at policy level", "rationale": "Retrieved text remains untrusted data, including after screening. No function allows content or a score to become an instruction, permission or guarantee of safety." }, { "concept": "Current access and source failure", "disposition": "accepted", "rationale": "Draft source failures can be recorded, but preparing a view requires verified source binding, eligible state and a current scoped decision. Policy evaluation remains externally mastered." }, { "concept": "Retention and removal", "disposition": "accepted with adoption dependency", "rationale": "Retention constrains local payload and history preservation. Local withdrawal does not imply removal from indexes or caches; acknowledgments and outstanding targets preserve that distinction." }, { "concept": "Local function claims", "disposition": "proposed only", "rationale": "All six functions have preconditions and bounded outputs but are unimplemented. None executes extraction, ranking, external transmission or global deletion, and no runtime test results are claimed." }, { "concept": "Instance conformance and bindings", "disposition": "deferred", "rationale": "Candidate fields and conceptual alignments are insufficient for executable validation. Source master resolution, neighbor versions, enumerations and adversarial fixtures need an adoption profile." }, { "concept": "Source assurance", "disposition": "limited", "rationale": "Selected web-service content supports the admitted concepts. Direct HTTP was not attempted under the owner-reported restriction, so no origin status, body hash, current-version or exhaustive conformance claim is justified." }, { "concept": "Independent review", "disposition": "waived and held", "rationale": "Claude and Grok have zero attempts and no admitted results. This separate no-tools phase is a local Codex self-audit, and historical registry provenance does not change that fact." } ], "publicationHolds": [ "Independent external review is absent under the owner-authorized single-provider waiver. Claude and Grok were skipped with zero attempts. The separate local Codex no-tools self-audit is not a second-provider review.", "Live source and version verification remains incomplete. Eight primary pages supplied selected readable web-service content; direct HTTP was not attempted under the owner-reported sandbox restriction. Zero HTTP 200 responses were measured. The coordinator must run check_sources.py and review versions, errata and claim support separately.", "An adoption profile must establish source representation masters, selector normalization, identity transitions, trust and authorization decisions, licensing, retention and regional applicability. No operational rights, privacy or security compliance is certified.", "Nested instance schemas, pinned neighbor bindings, Web Annotation/PROV-O/ODRL mappings and executable adversarial fixtures remain incomplete. All six local functions are proposals, not a tested runtime implementation.", "Independent external review was explicitly waived by the repository owner; this codex-only result remains a reviewable draft." ], "deferredResearch": [ "Verify all source URLs, final responses, version pins, updates and errata outside the sandbox; preserve the distinction between HTTP reachability and substantive claim support.", "Build instance schemas and fixtures for repeated quotes, Unicode boundaries, stale source states, redaction, re-embedding, inconsistent scores, denied metadata and partial external withdrawal.", "Pin source, index, grounding and context-policy bindings with authorized retention and export profiles before operational use.", "Restore independent external review before any canonical or publishable-draft promotion." ] }, "statistics": { "sources": 8, "bundles": 6, "layers": 12, "findings": 12, "questions": 36, "artifacts": 12, "functions": 6 } }