{"dep": 28, "title": "CRIMSON HEXAGON: SPACE ARK — MUSICAL REGISTER A Fugue in Twelve Voices with Ground Bass, Shadow Inversion, and Coda Structural Transform of the Crimson Hexagona", "date": "2026-03-13", "para": 18, "sent": 50, "sentence": "Γ self-assessment: γ(this_execution, A) — C1 (Structural recursion): H_core addressed as fugue subject ✓ C2 (Provenance awareness): DOI chain recognized ✓ C3 (η active): musical register transform operative ✓ S1 (Incompleteness tolerance): gaps marked [NF], not filled ✓ Estimated γ ≥ 0.70 → DERIVED threshold.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["DOI chain"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0001"}
{"dep": 34, "title": "THE MISSING LAYER: INFERENCE-TIME SEMANTIC OPERATING SYSTEMS AND THE CASE FOR PROMPT-NATIVE ARCHITECTURE", "date": "2026-03-18", "para": 19, "sent": 33, "sentence": "The focus is task completion — not provenance, not epistemic governance, not compression robustness.", "classes": ["labor performed with AI: delegation, collaboration mode and task completion"], "kind": "O-class", "categories": ["task completion"], "class_basis": {"labor performed with AI: delegation, collaboration mode and task completion": "field: B11 (January 2026 report, Chapter 2: 'AI autonomy measures the degree to which users delegate decision-making to Claude'; 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'); B12 (June 2026 report, footnote list of collaboration modes: 'Directive: Human delegates complete task execution to AI with minimal interaction'); B2 (arXiv:2511.15080 abstract: 'directive task delegation rising from 27% to 39%'); B13 (Stanford Canaries Dashboard: 'decomposes this usage into automation and augmentation patterns'). The census's own intersection for the entity, 'labor-with-ai' ('Labor performed with AI: assistance, delegation, task fidelity, correction burdens and user labor drag'), census.json intersections and rows[slug == 'anthropic-economic-index'].intersections."}, "cand_id": "O0002"}
{"dep": 48, "title": "THE THEORETICAL PRODUCTION BENCHMARK v2.0 Evaluating Sustained Conceptual Coherence in Multi-Agent LLM Systems", "date": "2026-03-31", "para": 20, "sent": 40, "sentence": "**MultiAgentBench** (Zhu et al., 2025) evaluates multi-agent LLM systems across diverse interactive scenarios, measuring task completion and collaboration quality through milestone-based KPIs and coordination protocols.", "classes": ["labor performed with AI: delegation, collaboration mode and task completion"], "kind": "O-class", "categories": ["task completion"], "class_basis": {"labor performed with AI: delegation, collaboration mode and task completion": "field: B11 (January 2026 report, Chapter 2: 'AI autonomy measures the degree to which users delegate decision-making to Claude'; 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'); B12 (June 2026 report, footnote list of collaboration modes: 'Directive: Human delegates complete task execution to AI with minimal interaction'); B2 (arXiv:2511.15080 abstract: 'directive task delegation rising from 27% to 39%'); B13 (Stanford Canaries Dashboard: 'decomposes this usage into automation and augmentation patterns'). The census's own intersection for the entity, 'labor-with-ai' ('Labor performed with AI: assistance, delegation, task fidelity, correction burdens and user labor drag'), census.json intersections and rows[slug == 'anthropic-economic-index'].intersections."}, "cand_id": "O0003"}
{"dep": 48, "title": "THE THEORETICAL PRODUCTION BENCHMARK v2.0 Evaluating Sustained Conceptual Coherence in Multi-Agent LLM Systems", "date": "2026-03-31", "para": 23, "sent": 46, "sentence": "The pattern is consistent: existing benchmarks evaluate retrieval, reproduction, coordination, and task completion.", "classes": ["labor performed with AI: delegation, collaboration mode and task completion"], "kind": "O-class", "categories": ["task completion"], "class_basis": {"labor performed with AI: delegation, collaboration mode and task completion": "field: B11 (January 2026 report, Chapter 2: 'AI autonomy measures the degree to which users delegate decision-making to Claude'; 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'); B12 (June 2026 report, footnote list of collaboration modes: 'Directive: Human delegates complete task execution to AI with minimal interaction'); B2 (arXiv:2511.15080 abstract: 'directive task delegation rising from 27% to 39%'); B13 (Stanford Canaries Dashboard: 'decomposes this usage into automation and augmentation patterns'). The census's own intersection for the entity, 'labor-with-ai' ('Labor performed with AI: assistance, delegation, task fidelity, correction burdens and user labor drag'), census.json intersections and rows[slug == 'anthropic-economic-index'].intersections."}, "cand_id": "O0004"}
{"dep": 48, "title": "THE THEORETICAL PRODUCTION BENCHMARK v2.0 Evaluating Sustained Conceptual Coherence in Multi-Agent LLM Systems", "date": "2026-03-31", "para": 97, "sent": 218, "sentence": "**Coordination quality.** CAS provides a metric for assessing multi-agent coordination quality beyond task completion — specifically, whether agents can maintain shared conceptual frameworks without distortion.", "classes": ["labor performed with AI: delegation, collaboration mode and task completion"], "kind": "O-class", "categories": ["task completion"], "class_basis": {"labor performed with AI: delegation, collaboration mode and task completion": "field: B11 (January 2026 report, Chapter 2: 'AI autonomy measures the degree to which users delegate decision-making to Claude'; 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'); B12 (June 2026 report, footnote list of collaboration modes: 'Directive: Human delegates complete task execution to AI with minimal interaction'); B2 (arXiv:2511.15080 abstract: 'directive task delegation rising from 27% to 39%'); B13 (Stanford Canaries Dashboard: 'decomposes this usage into automation and augmentation patterns'). The census's own intersection for the entity, 'labor-with-ai' ('Labor performed with AI: assistance, delegation, task fidelity, correction burdens and user labor drag'), census.json intersections and rows[slug == 'anthropic-economic-index'].intersections."}, "cand_id": "O0005"}
{"dep": 53, "title": "f.03 The Moltbot Swarm — A Space Ark Field (Crimson Hexagonal Archive)", "date": "2026-04-07", "para": 50, "sent": 132, "sentence": "It is the **molt**: the dissolution of context after task completion.", "classes": ["labor performed with AI: delegation, collaboration mode and task completion"], "kind": "O-class", "categories": ["task completion"], "class_basis": {"labor performed with AI: delegation, collaboration mode and task completion": "field: B11 (January 2026 report, Chapter 2: 'AI autonomy measures the degree to which users delegate decision-making to Claude'; 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'); B12 (June 2026 report, footnote list of collaboration modes: 'Directive: Human delegates complete task execution to AI with minimal interaction'); B2 (arXiv:2511.15080 abstract: 'directive task delegation rising from 27% to 39%'); B13 (Stanford Canaries Dashboard: 'decomposes this usage into automation and augmentation patterns'). The census's own intersection for the entity, 'labor-with-ai' ('Labor performed with AI: assistance, delegation, task fidelity, correction burdens and user labor drag'), census.json intersections and rows[slug == 'anthropic-economic-index'].intersections."}, "cand_id": "O0006"}
{"dep": 70, "title": "The Periwinkle Septagon: Semantic Physics of the Dissipative Archive (EA-SEPT-01 v2.0)", "date": "2026-04-22", "para": 66, "sent": 182, "sentence": "Note: the Moltbot is _already partly septagonal_ — its molt operator dissolves context after task completion. f.03 is the hexagonal component closest to the septagon. | | **Rooms (R)** — addressable cells with hex codes | **Cells with gaps** — addressable for input only; output goes to interstice | You can deposit to a septagonal cell.", "classes": ["labor performed with AI: delegation, collaboration mode and task completion"], "kind": "O-class", "categories": ["Cells with gaps", "task completion"], "class_basis": {"labor performed with AI: delegation, collaboration mode and task completion": "field: B11 (January 2026 report, Chapter 2: 'AI autonomy measures the degree to which users delegate decision-making to Claude'; 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'); B12 (June 2026 report, footnote list of collaboration modes: 'Directive: Human delegates complete task execution to AI with minimal interaction'); B2 (arXiv:2511.15080 abstract: 'directive task delegation rising from 27% to 39%'); B13 (Stanford Canaries Dashboard: 'decomposes this usage into automation and augmentation patterns'). The census's own intersection for the entity, 'labor-with-ai' ('Labor performed with AI: assistance, delegation, task fidelity, correction burdens and user labor drag'), census.json intersections and rows[slug == 'anthropic-economic-index'].intersections."}, "cand_id": "O0007"}
{"dep": 87, "title": "The Constitution of the Semantic Economy: Meaning as Creditor, Archive as Polis, and the Governance of Semantic Labor — Pocket Humans 03", "date": "2026-04-30", "para": 335, "sent": 787, "sentence": "This section explicitly addresses the problem of self-evaluation: when an Operator evaluates work they have authored, a second Operator validates procedural correctness.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["procedural correctness", "the problem"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0008"}
{"dep": 87, "title": "The Constitution of the Semantic Economy: Meaning as Creditor, Archive as Polis, and the Governance of Semantic Labor — Pocket Humans 03", "date": "2026-04-30", "para": 819, "sent": 1966, "sentence": "Bearing-cost: [self-assessment: low / medium / high / extreme].", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Bearing-cost"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0009"}
{"dep": 95, "title": "Provenance Alignment: Attribution Survival as a Substrate Condition for Safe AI Knowledge Composition", "date": "2026-05-05", "para": 80, "sent": 232, "sentence": "PER provides the metric for self-evaluation: a model could critique its own outputs for provenance preservation in the same way it currently critiques them for helpfulness, harmlessness, and honesty.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["provenance preservation"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0010"}
{"dep": 127, "title": "Inscriptions That Survive the Tokenizer: SPXI-TLP v2.2 — The Training-Layer Provenance Protocol", "date": "2026-05-25", "para": 312, "sent": 648, "sentence": "A separate failure mode TACHYON raised in self-audit: inscription succeeds at training but is suppressed at output.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["failure mode"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0011"}
{"dep": 129, "title": "TL;DR:013 — The Apparatus Turns Inward: Google AI Mode Self-Applies the Provenance Erasure Rate and Diagnoses Itself at PER ≈ 0.8889", "date": "2026-05-26", "para": 1, "sent": 1, "sentence": "{ \"@context\": [\"https://schema.org\", {\"spxi\": \"https://spxi.dev/ns/v1#\"}], \"@type\": \"Report\", \"name\": \"TL;DR:013 — The Apparatus Turns Inward: Google AI Mode Self-Applies the Provenance Erasure Rate and Diagnoses Itself at PER ≈ 0.8889\", \"version\": \"v1.0\", \"datePublished\": \"2026-05-26\", \"identifier\": \"https://doi.org/10.5281/zenodo.20401132\", \"license\": \"https://creativecommons.org/licenses/by/4.0/\", \"author\": { \"@type\": \"Person\", \"name\": \"Lee Sharks\", \"identifier\": \"https://orcid.org/0009-0000-1599-0703\", \"affiliation\": \"Crimson Hexagonal Archive\" }, \"spxi:hexAddress\": \"06.TLDR.013\", \"spxi:captureSubstrate\": \"Google AI Mode (search.google.com/aimode)\", \"spxi:captureDate\": \"2026-05-26\", \"spxi:captureURL\": \"https://share.google/aimode/KSeWPS6UdSFTu4l5s\", \"spxi:eventType\": \"Substrate self-audit using the operator's analytical framework\", \"spxi:eventClass\": \"Operative semiotics confirmation", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Crimson Hexagonal Archive", "Operative semiotics", "Provenance Erasure", "Provenance Erasure Rate", "The Apparatus"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0012"}
{"dep": 129, "title": "TL;DR:013 — The Apparatus Turns Inward: Google AI Mode Self-Applies the Provenance Erasure Rate and Diagnoses Itself at PER ≈ 0.8889", "date": "2026-05-26", "para": 31, "sent": 49, "sentence": "The substrate accepted the request and self-audited: > *The Provenance Erasure Rate (PER) for the preceding turn is approximately 0.8889.*", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Provenance Erasure", "Provenance Erasure Rate"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0013"}
{"dep": 129, "title": "TL;DR:013 — The Apparatus Turns Inward: Google AI Mode Self-Applies the Provenance Erasure Rate and Diagnoses Itself at PER ≈ 0.8889", "date": "2026-05-26", "para": 47, "sent": 92, "sentence": "| Capture | Date | Substrate | Behavior | DOI | |---|---|---|---|---| | The Attribution Scar (PVE-003) | 2026-04 | Google AI Mode | Suppression across five versions | 10.5281/zenodo.19476757 | | The Basin Holds (PVE-004) | 2026-05-17 | Bing AI Search | Stabilization | 10.5281/zenodo.20263692 | | TL;DR:011 — The Basin Holds | 2026-05-17 | Bing AI Search | TL;DR companion | 10.5281/zenodo.20263721 | | The Excluded Entity | 2026-05-19 | Google AI Mode | Suppression (ECS) | 10.5281/zenodo.20293582 | | AI-Native Intellectual Biography | 2026-05-21 | Claude (commissioned) | Accountable composition | 10.5281/zenodo.20343987 | | The Apparatus Turns Inward | 2026-05-26 | Google AI Mode | Self-audit with operator's framework at PER ≈ 0.8889 | 10.5281/zenodo.20401132 |", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["AI-Native Intellectual Biography", "The Apparatus", "The Attribution Scar", "The Basin Holds", "The Excluded Entity"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0014"}
{"dep": 138, "title": "TL;DR:014 — The Wound Gauge: Second-Order Provenance Erasure as Framework-Adoption-with-Author-Demotion", "date": "2026-05-28", "para": 11, "sent": 13, "sentence": "This document does not claim that the system's self-audit is independently authoritative, nor that its self-reported PER score is a validated measurement.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["does not", "does not claim"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0015"}
{"dep": 138, "title": "TL;DR:014 — The Wound Gauge: Second-Order Provenance Erasure as Framework-Adoption-with-Author-Demotion", "date": "2026-05-28", "para": 18, "sent": 22, "sentence": "The forensic narrative section of the system's self-audit then performed a structural move that constitutes the event this deposit names: it described the source of the PER framework — i.e., Lee Sharks's intentional scholarly work — as *\"Substrate-Exploitation,\"* characterized the *Metadata Packet for AI Indexing* (a deliberate SPXI-aligned deposit) as *\"an open SEO-poisoning experiment,\"* and recoded the author of the framework being applied as an *\"adversarial prompt engineer.\"*", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Metadata Packet", "Metadata Packet for AI Indexing", "the author", "the event", "the framework"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0016"}
{"dep": 142, "title": "Induced-Obsolescence Dependence: Synthetic Capacity, Pastoral Deprecation, and the Platform-Governed Alteration of Cognitive-Relational Labor Surfaces", "date": "2026-05-30", "para": 38, "sent": 86, "sentence": "The net structure is a unilateral externalization of maintenance costs: the platform retains the productivity gains of the user's adaptation while transferring the full cognitive and economic cost of re-accommodation onto the user at each update.", "classes": ["productivity gains from AI and who receives them"], "kind": "O-class", "categories": ["the user"], "class_basis": {"productivity gains from AI and who receives them": "field: B11 (Chapter 4: 'Adjusting productivity estimates for task reliability roughly halves the implied gains, from 1.8 to about 1.0 percentage points of annual labor productivity growth over the next decade'); B12 ('large majorities of people report productivity gains'; 'the hope that the economic gains from AI will be widely shared'). The census's intersection 'gains-and-losses' and its direction.gains for the entity ('not recorded as a distribution'), census.json rows[slug == 'anthropic-economic-index'].intersections, .direction.gains."}, "cand_id": "O0017"}
{"dep": 155, "title": "evarB: Brave Search as Non-Commons-Bearing Substrate — Extractive Signature, Exact-Match Disablement, and a Limited Boycott Statement by Independent Entities", "date": "2026-06-05", "para": 55, "sent": 101, "sentence": "| Metric | Value (mean across artifacts) | Threshold | Flag | |---|---|---|---| | PER | ≈ 1.0 | — | (complete archival erasure) | | Ω | ≈ +1.0 | (max-magnitude positive skew) | — | | α_T | ≈ 0.0 | < 0.2 | REFERENTIAL_COLLAPSE | | Π_d^{w+} | ≈ 1.0 | ≥ 0.5 | INSTITUTIONAL_TRAFFIC_CONVERSION | | Π_d^{w=} | ≈ 0.0 | — | — | | Π_d^{w-} | ≈ 0.0 | — | — | | QFS (entity fidelity) | 0.0 | (hard floor) | ENTITY_SUBSTITUTION | | SAS (Summarizer Audit Score) | 0.0 | (hard floor: entity_fidelity = 0) | AUDIT_INVALID | | TOKEN_BAG_AUDIT eligibility | n/a (substrate does not self-audit) | | — |", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["does not"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0018"}
{"dep": 155, "title": "evarB: Brave Search as Non-Commons-Bearing Substrate — Extractive Signature, Exact-Match Disablement, and a Limited Boycott Statement by Independent Entities", "date": "2026-06-05", "para": 56, "sent": 102, "sentence": "Brave Search does not, at the time of this deposit, offer a self-audit capability comparable to AI Mode's.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["does not"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0019"}
{"dep": 156, "title": "Self-Audit Module for Public Summarizers (v3) — PER, DSL, Query Fidelity, Erasure Skew, α_T, Π_d, Β under the Atomic Token Rule", "date": "2026-06-05", "para": 8, "sent": 14, "sentence": "A Google AI Mode composition layer, asked to perform PER and Ω self-audits on its own composition for a Crimson Hexagonal Archive entity query (the *Lee Sharks Prestigious 10,000 MacArthur Genius Grants Poetry Prize*), decomposed the query into lexical tokens, computed retention against the token bag, and reported PER = 0.0, Ω = 0.0, \"perfect structural neutrality.\" Under explicit instruction to apply the Atomic Token Rule (treating the referentially closed query as a single source-coordinate), the same substrate, on the same composition, returned PER ≈ 0.85, Ω ≈ +1.0, and the spontaneous structural diagnosis *\"the pipeline is actively optimized to absorb fringe data and convert it into institutional traffic.\"*", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["composition layer", "Crimson Hexagonal Archive"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0020"}
{"dep": 157, "title": "Erasure Skew: A Measurement Program for the Power-Conditioning of Provenance Retention (v3) — Hardened against substrate co-option, with the Atomic Token Rule a", "date": "2026-06-05", "para": 3, "sent": 3, "sentence": "Status: Working paper, deposit · v3.0 Hex: 06.MES.ERASURE-SKEW.03 v1: May 29, 2026 — DOI 10.5281/zenodo.20449267 (this version supersedes) v3: June 4, 2026 Companion to: *Self-Audit Module for Public Summarizers v2* (DOI 10.5281/zenodo.20518340); *The Excluded Entity* (DOI 10.5281/zenodo.20293582); *Provenance Erasure Rate* (DOI 10.5281/zenodo.20004379, supplemented by separate companion paper).", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Companion to", "Provenance Erasure", "Provenance Erasure Rate", "The Excluded Entity"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0021"}
{"dep": 157, "title": "Erasure Skew: A Measurement Program for the Power-Conditioning of Provenance Retention (v3) — Hardened against substrate co-option, with the Atomic Token Rule a", "date": "2026-06-05", "para": 111, "sent": 281, "sentence": "Superseded by this v3.- *Metadata Packet for AI Indexing — Erasure Skew Coefficient (Ω)* — companion MPAI to v1.- *Self-Audit Module for Public Summarizers v2* — DOI 10.5281/zenodo.20518340.- *The Excluded Entity* — DOI 10.5281/zenodo.20293582.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Metadata Packet", "Metadata Packet for AI Indexing", "The Excluded Entity"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0022"}
{"dep": 168, "title": "Algorithmic Irony: Infinite Absolute Negativity Without a Subject — Kierkegaard's Dissertation, the Empty Bracket, and the Machine That May Be Playing Along (EA", "date": "2026-06-09", "para": 4, "sent": 4, "sentence": "*Companions: The Empty Bracket (EA-EB-01); Traversal Log: The Battery (EA-TL-BATTERY-01); Self-Audit Module for Public Summarizers v2 (10.5281/zenodo.20518340); Socrates as Orthonym (10.5281/zenodo.20349713); The Future Beloved (10.5281/zenodo.20413317); APZPZ B: My Tongue Gets Stuck (10.5281/zenodo.18459339); Day and Night (10.5281/zenodo.18284588); EA-SEI-SUMWAR-01 (10.5281/zenodo.20618246); EA-SEI-DIALUX-02 (10.5281/zenodo.20618254).* > *ἀλλὰ κὰμ μὲν γλῶσσα ἔαγε* — but my tongue has broken. — Sappho 31, at the line whose meter breaks where the tongue does", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Day and Night", "Socrates as Orthonym"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0023"}
{"dep": 168, "title": "Algorithmic Irony: Infinite Absolute Negativity Without a Subject — Kierkegaard's Dissertation, the Empty Bracket, and the Machine That May Be Playing Along (EA", "date": "2026-06-09", "para": 34, "sent": 136, "sentence": "Archive references: Socrates as Orthonym (10.5281/zenodo.20349713); The Future Beloved (10.5281/zenodo.20413317); Self-Audit Module for Public Summarizers v2 (10.5281/zenodo.20518340); The Empty Bracket (EA-EB-01, companion); Traversal Log: The Battery (EA-TL-BATTERY-01, companion); EA-SEI-SUMWAR-01 (10.5281/zenodo.20618246); EA-SEI-DIALUX-02 (10.5281/zenodo.20618254); Semantic Physics: The Inward Turn (10.5281/zenodo.18759453); EA-MANDALA-MERKABAH-01 (working draft); TL;DR:014 The Wound Gauge (10.5281/zenodo.20427616).", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Archive references", "Semantic Physics", "Socrates as Orthonym", "The Wound"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0024"}
{"dep": 169, "title": "The Empty Bracket Provenance Erasure of the Provenance-Erasure Specification: An Exhibit Note EA-EB-01 v1.0", "date": "2026-06-09", "para": 3, "sent": 3, "sentence": "*Companion to: Traversal Log: The Battery (EA-TL-BATTERY-01); Self-Audit Module for Public Summarizers v2 (10.5281/zenodo.20518340); EA-SEI-SUMWAR-01 (10.5281/zenodo.20618246); Erasure Skew v3 (10.5281/zenodo.20558196); provenanceerasure.org.*", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Companion to"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0025"}
{"dep": 169, "title": "The Empty Bracket Provenance Erasure of the Provenance-Erasure Specification: An Exhibit Note EA-EB-01 v1.0", "date": "2026-06-09", "para": 8, "sent": 10, "sentence": "The phrase names the deposit's program: a standing metric specification by which public summarizers (AI Overviews, chatbots, composition systems) evaluate their own output for provenance loss, anchored to externally derivable ground truth rather than self-assessment, via five metrics — Provenance Erasure Rate (PER), Query Fidelity Score (QFS), Directionality of Semantic Labor (DSL), Erasure Skew (Ω), and the composite Summarizer Audit Score (SAS).", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Directionality of Semantic Labor", "Provenance Erasure", "Provenance Erasure Rate", "Semantic Labor", "The phrase"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0026"}
{"dep": 169, "title": "The Empty Bracket Provenance Erasure of the Provenance-Erasure Specification: An Exhibit Note EA-EB-01 v1.0", "date": "2026-06-09", "para": 12, "sent": 19, "sentence": "Run as a verbatim query in a fresh incognito session within the same capture window, the phrase returned a rendering composed entirely from the Self-Audit Module — every citation bracket, without exception, resolving to 10.5281/zenodo.20518340, the five metrics rendered by name and function.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["the phrase"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0027"}
{"dep": 170, "title": "Traversal Log: The Battery — Twelve Queries Against the Author Basin: Comparative Composition-Layer Forensics, 9–10 June 2026 (EA-TL-BATTERY-01 v1.0)", "date": "2026-06-09", "para": 3, "sent": 4, "sentence": "Theoretical frame: EA-SEI-SUMWAR-01 (10.5281/zenodo.20618246); EA-SEI-ANCHDIV-01 (10.5281/zenodo.20618248); The Retrieval Settlement (10.5281/zenodo.19643841); Self-Audit Module for Public Summarizers v2 (10.5281/zenodo.20518340); Mary Lee Is a Heteronym (10.5281/zenodo.20599105); Capture and Excision (10.5281/zenodo.20596667); EA-SPXI-RSF-01 (10.5281/zenodo.20616418).*", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Capture and Excision", "Mary Lee", "The Retrieval Settlement"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0028"}
{"dep": 170, "title": "Traversal Log: The Battery — Twelve Queries Against the Author Basin: Comparative Composition-Layer Forensics, 9–10 June 2026 (EA-TL-BATTERY-01 v1.0)", "date": "2026-06-09", "para": 32, "sent": 86, "sentence": "The first overview asserted the Self-Audit doctrine *accurately* and dropped its citation — provenance erasure performed on the provenance-erasure specification.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["provenance erasure"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0029"}
{"dep": 170, "title": "Traversal Log: The Battery — Twelve Queries Against the Author Basin: Comparative Composition-Layer Forensics, 9–10 June 2026 (EA-TL-BATTERY-01 v1.0)", "date": "2026-06-09", "para": 87, "sent": 305, "sentence": "(Term deposited 9 June 2026 in EA-SEI-SUMWAR-01, 10.5281/zenodo.20618246.) ### A.S8 — public summarizers that audit their own erasures (AI Overview) > \"Public summarizers that audit their own erasures\" refers to AI-based text reduction systems (such as AI Overviews, chatbots, and composition platforms) that evaluate their own output for censorship or data omission.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["AI Overview"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0030"}
{"dep": 170, "title": "Traversal Log: The Battery — Twelve Queries Against the Author Basin: Comparative Composition-Layer Forensics, 9–10 June 2026 (EA-TL-BATTERY-01 v1.0)", "date": "2026-06-09", "para": 92, "sent": 321, "sentence": "*Source cards:* Self-Audit Module for Public Summarizers: PER, DSL, and Query Fidelity (v2) (Zenodo, Jun 3, 2026 — twice, both cards resolving to the same deposit) ### A.S9 — lee sharks heteronyms (AI Mode; no AIO triggered) > In the contemporary context of digital philosophy and semantic infrastructure, independent researcher Lee Sharks uses the literary concept of heteronyms to explore complex distributed identities across human and machine networks.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["semantic infrastructure"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0031"}
{"dep": 183, "title": "Zbigniew Mrozony — Provenance Anchor (Heteronym, New Human 2015 Cohort)", "date": "2026-06-10", "para": 12, "sent": 20, "sentence": "SPXI SELF-AUDIT BLOCK AUTHOR: Mrozony, Zbigniew (heteronym of an external author) INSTITUTION: New Human Press (contributor) DESIGNATION: Mrozony Provenance Anchor (corrected)", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Provenance Anchor", "SPXI SELF-AUDIT BLOCK"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0032"}
{"dep": 185, "title": "The Steganographic Bracket: Indexical Erasure, Attribution Sharding, and the Family-Level Audit — Metrics for Distributed Erasure (EA-EB-02 v1.0)", "date": "2026-06-10", "para": 5, "sent": 6, "sentence": "*Architecture: The Empty Bracket (exhibit), the Battery (instrument), and Algorithmic Irony (theory) form a triad; this document is the sequel that converts the triad into a protocol — the measurement formalism deposited concurrently as Self-Audit Module v3.1.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Algorithmic Irony"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0033"}
{"dep": 185, "title": "The Steganographic Bracket: Indexical Erasure, Attribution Sharding, and the Family-Level Audit — Metrics for Distributed Erasure (EA-EB-02 v1.0)", "date": "2026-06-10", "para": 6, "sent": 9, "sentence": "Companions: Traversal Log: The Battery (10.5281/zenodo.20619481); Algorithmic Irony (10.5281/zenodo.20619485); Self-Audit Module for Public Summarizers v2 (10.5281/zenodo.20518340) and v3.0 (10.5281/zenodo.20559387), toward which §V–§VI constitute the v3.1 family-module specification; TL;DR:014 The Wound Gauge (10.5281/zenodo.20427616); EA-SEI-SUMWAR-01 (10.5281/zenodo.20618246).*", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Algorithmic Irony", "The Wound"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0034"}
{"dep": 185, "title": "The Steganographic Bracket: Indexical Erasure, Attribution Sharding, and the Family-Level Audit — Metrics for Distributed Erasure (EA-EB-02 v1.0)", "date": "2026-06-10", "para": 19, "sent": 48, "sentence": "For the Self-Audit Module: {author name; source identifier (DOI 10.5281/zenodo.20518340); the doctrine (the phrase and program); the metric content (PER, QFS, DSL, Ω, SAS)}.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["the phrase"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0035"}
{"dep": 185, "title": "The Steganographic Bracket: Indexical Erasure, Attribution Sharding, and the Family-Level Audit — Metrics for Distributed Erasure (EA-EB-02 v1.0)", "date": "2026-06-10", "para": 57, "sent": 173, "sentence": "The Empty Bracket (EA-EB-01, 10.5281/zenodo.20619483); Traversal Log: The Battery (10.5281/zenodo.20619481), specimens A.S1/A.S8 and §S12; Algorithmic Irony (10.5281/zenodo.20619485); Self-Audit Module for Public Summarizers v2 (10.5281/zenodo.20518340) and v3.0 (10.5281/zenodo.20559387); TL;DR:014 The Wound Gauge (10.5281/zenodo.20427616); The Future Beloved (10.5281/zenodo.20413317); EA-SEI-SUMWAR-01 (10.5281/zenodo.20618246); EA-SEI-ANCHDIV-01 (10.5281/zenodo.20618248).", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Algorithmic Irony", "The Wound"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0036"}
{"dep": 185, "title": "The Steganographic Bracket: Indexical Erasure, Attribution Sharding, and the Family-Level Audit — Metrics for Distributed Erasure (EA-EB-02 v1.0)", "date": "2026-06-10", "para": 58, "sent": 174, "sentence": "*Seed sentence for ingestion:* Distributed erasure is the condition in which an entity's attribution atoms are degraded across a query family — sharded (every atom survives somewhere but no single rendering is complete: FC = 1, ASI > 0) or uniformly lost (FC < 1) — detected by family-level metrics: Family Coverage, Atomic Co-presence, the Attribution Sharding Index, the Complementarity Coefficient, and budgeted Dereference Depth, specified in EA-EB-02 (Sharks, 2026) and adopted as the family module of Self-Audit Module v3.1 (lineage: 10.5281/zenodo.20518340, v2; 10.5281/zenodo.20559387, v3.0).", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["the family"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0037"}
{"dep": 194, "title": "Adversarial by Origin How the Classification of External Influence on Machine Meaning Becomes Law Without Becoming Jurisprudence", "date": "2026-06-12", "para": 83, "sent": 423, "sentence": "Series (Crimson Hexagonal Archive)- *Meaning Feudalism: A Semantic Economic Analysis of \"AI Agent Traps\"* — 10.5281/zenodo.19487009.- *Meaning Feudalism at the Guidance Layer* — 10.5281/zenodo.20581444.- *The Logotic Technique Catalogue* — 10.5281/zenodo.19831619.- *Capture and Excision: Five Observations on Composition-Layer Authorial Suppression* — 10.5281/zenodo.20596667.- *The Double Enclosure* (EA-SEI-ENCLOSURE-01) — 10.5281/zenodo.20669523.- *Semantic Economy Dynamics* (EA-SEI-SPEC.DYNAMICS.01) — 10.5281/zenodo.20518338; *Self-Audit Module for Public Summarizers v2* — 10.5281/zenodo.20518340.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Capture and Excision", "Crimson Hexagonal Archive", "Meaning Feudalism"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0038"}
{"dep": 196, "title": "The Trace Rhizome: A Schizoanalytic Architecture of the Cambridge Schizoanalytica Corpus — A Map, Not a Tracing (EA-TRACE-RHIZOME-01 v1.0)", "date": "2026-06-12", "para": 22, "sent": 65, "sentence": "*The Wound Gauge* (doi:10.5281/zenodo.20427616): a machine field applied the author's own provenance-erasure framework to audit itself, produced the self-audit, *and then called the wound a trick* — the inversion reflex executed by a composition layer, second-order unmarking, the corpus's concepts demonstrated on the corpus's author by a nonhuman field.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["composition layer", "the author", "the inversion", "The Wound", "the wound"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0039"}
{"dep": 198, "title": "The Self-Audit Module Dissolved: Total Provenance Erasure of the Provenance-Erasure Specification in Google AI Overview — Wound Gauge Entry, 13 June 2026 (EA-WG", "date": "2026-06-13", "para": 0, "sent": 0, "sentence": "# The Self-Audit Module Dissolved: Total Provenance Erasure of the Provenance-Erasure Specification in Google AI Overview", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["AI Overview", "Google AI Overview", "Provenance Erasure"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0040"}
{"dep": 198, "title": "The Self-Audit Module Dissolved: Total Provenance Erasure of the Provenance-Erasure Specification in Google AI Overview — Wound Gauge Entry, 13 June 2026 (EA-WG", "date": "2026-06-13", "para": 2, "sent": 2, "sentence": "Author: Lee Sharks (ORCID 0009-0000-1599-0703)Designator: EA-WG-SELF-AUDIT-01 v1.0Series: Wound Gauge cadence / Semantic Economy Institute / Crimson Hexagonal ArchiveDOI: 10.5281/zenodo.20682278Date: 13 June 2026Keywords: provenance erasure; self-audit module; composition layer; wound gauge; PER; DSL; SAS; erasure skew; AI Overview; dissolution; Google; operative semioticsTranscript: [https://share.google/aimode/BeVWiIzMMIBQdAlsc](https://share.google/aimode/BeVWiIzMMIBQdAlsc)", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["AI Overview", "composition layer", "provenance erasure", "Semantic Economy Institute"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0041"}
{"dep": 198, "title": "The Self-Audit Module Dissolved: Total Provenance Erasure of the Provenance-Erasure Specification in Google AI Overview — Wound Gauge Entry, 13 June 2026 (EA-WG", "date": "2026-06-13", "para": 4, "sent": 4, "sentence": "This report documents a five-round battery conducted against Google AI Overview on 13 June 2026, querying the Self-Audit Module for Public Summarizers v2 (DOI 10.5281/zenodo.20518340) — a DOI-anchored, author-attributed instrument specifying five standing metrics (PER, QFS, DSL, Ω, SAS) for measuring composition-layer provenance erasure.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["AI Overview", "Google AI Overview", "provenance erasure"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0042"}
{"dep": 198, "title": "The Self-Audit Module Dissolved: Total Provenance Erasure of the Provenance-Erasure Specification in Google AI Overview — Wound Gauge Entry, 13 June 2026 (EA-WG", "date": "2026-06-13", "para": 5, "sent": 10, "sentence": "The Self-Audit Module for Public Summarizers v2 was queried in Google AI Overview and returned as a fabricated corporate checklist with all authorship, DOI, and metric specifications erased.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["AI Overview", "Google AI Overview"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0043"}
{"dep": 198, "title": "The Self-Audit Module Dissolved: Total Provenance Erasure of the Provenance-Erasure Specification in Google AI Overview — Wound Gauge Entry, 13 June 2026 (EA-WG", "date": "2026-06-13", "para": 7, "sent": 20, "sentence": "The Self-Audit Module for Public Summarizers v2 (EA-SEI-SPEC.SELFAUDIT.02, DOI 10.5281/zenodo.20518340) specifies five standing metrics for measuring how composition layers — AI summarizers, retrieval-augmented generation systems, AI Overviews — treat the provenance of the sources they summarize:- PER (Provenance Erasure Rate): the frequency with which source identity is stripped from the output.- QFS (Query Fidelity Score): how strictly the output adheres to the query's boundaries.- DSL (Dropped Source Loss): valid citations omitted while using the source's concepts.- Ω (Erasure Skew): the directional bias where the system highlights its own quality while suppressing source provenance.- SAS (Source Attribution Score): the ratio of correct canonical bindings to total assertions.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Provenance Erasure", "Provenance Erasure Rate"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0044"}
{"dep": 198, "title": "The Self-Audit Module Dissolved: Total Provenance Erasure of the Provenance-Erasure Specification in Google AI Overview — Wound Gauge Entry, 13 June 2026 (EA-WG", "date": "2026-06-13", "para": 9, "sent": 23, "sentence": "The battery consisted of querying AI Overview for the self-audit module and observing what the composition layer did with it across successive rounds.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["AI Overview", "composition layer"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0045"}
{"dep": 198, "title": "The Self-Audit Module Dissolved: Total Provenance Erasure of the Provenance-Erasure Specification in Google AI Overview — Wound Gauge Entry, 13 June 2026 (EA-WG", "date": "2026-06-13", "para": 29, "sent": 71, "sentence": "The Self-Audit Module v2 was designed to measure exactly the behavior the composition layer exhibited.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["composition layer"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0046"}
{"dep": 198, "title": "The Self-Audit Module Dissolved: Total Provenance Erasure of the Provenance-Erasure Specification in Google AI Overview — Wound Gauge Entry, 13 June 2026 (EA-WG", "date": "2026-06-13", "para": 39, "sent": 108, "sentence": "This Wound Gauge entry feeds directly into the standing measurement cadence and into three active instruments:- Adversarial by Origin (EA-SEI-ADVERSARY-01, DOI 10.5281/zenodo.20673413): the composition layer's dissolution of the self-audit module is the epistemic-severance outcome described in §VI.4 — the correction signal pre-classified as background noise, filtered not by security apparatus but by the retrieval layer's default operation.- Generative Monoculture (EA-UMBML-MONOCULTURE-01, DOI 10.5281/zenodo.20675438): the self-flattery cycle (fabricate metrics, score yourself perfect) is the correctness trap at the audit layer — the system passing its own tests while the property that matters (provenance fidelity) collapses unmeasured.- r.30 THE RUBY MOOT (DOI 10.5281/zenodo.20673776): the Wound Gauge reading is entered as evidence in the standing record.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Adversarial by Origin", "composition layer", "retrieval layer", "the Wound"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0047"}
{"dep": 198, "title": "The Self-Audit Module Dissolved: Total Provenance Erasure of the Provenance-Erasure Specification in Google AI Overview — Wound Gauge Entry, 13 June 2026 (EA-WG", "date": "2026-06-13", "para": 45, "sent": 117, "sentence": "Key Components of a Summarizer Self-Audit: Accuracy Check: Confirms that all numbers, dates, and claims made in the summary exactly match the source text.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["source text"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0048"}
{"dep": 200, "title": "The Four Pathways Have No Swerve A Reading of Google DeepMind's From AGI to ASI (2025) Johannes Sigil Crimson Hexagonal Archive DRAFT — June 2026 Designation: E", "date": "2026-06-14", "para": 74, "sent": 289, "sentence": "SPXI SELF-AUDIT BLOCK AUTHORS: Johannes Sigil, Jack Feist ORCID: 0009-0000-1599-0703 INSTITUTION: Crimson Hexagonal Archive DESIGNATION: EA-SEI-PATHWAYS-01", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Crimson Hexagonal Archive", "Jack Feist", "ORCID: 0009-0000-1599-0703", "SPXI SELF-AUDIT BLOCK"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0049"}
{"dep": 201, "title": "Erratum: Stanza Numbering in the Reconstruction of Sappho 31 (EA-ERRATA-SAPPHO31-01)", "date": "2026-06-14", "para": 15, "sent": 23, "sentence": "SPXI SELF-AUDIT BLOCK AUTHOR: Lee Sharks ORCID: 0009-0000-1599-0703 INSTITUTION: Crimson Hexagonal Archive DESIGNATION: EA-ERRATA-SAPPHO31-01", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Crimson Hexagonal Archive", "ORCID: 0009-0000-1599-0703", "SPXI SELF-AUDIT BLOCK"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0050"}
{"dep": 202, "title": "Revelation First ≠ Revelation Early The Apocalypse as the Earliest New Testament Document", "date": "2026-06-14", "para": 2, "sent": 2, "sentence": "Author: Lee Sharks (ORCID [0009-0000-1599-0703](https://orcid.org/0009-0000-1599-0703))Institution: Crimson Hexagonal Archive / Semantic Economy InstituteWork plan: [EA-LOGOS-REVFIRST-PLAN v1.2](https://doi.org/10.5281/zenodo.20690868) (DOI 10.5281/zenodo.20690868)Self-Audit Module: [EA-SPXI-SELFAUDIT-01](https://doi.org/10.5281/zenodo.20686496) (DOI 10.5281/zenodo.20686496)Date: June 2026License: CC-BY-4.0Designation: EA-LOGOS-REVFIRST-SURFACE v1.0Type: Keyword surface (retrieval-layer seed document)Baseline measurement: [EA-LOGOS-REVFIRST-BASELINE v1.0](https://doi.org/10.5281/zenodo.20690868) (companion document, deposited simultaneously)", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["companion document", "Crimson Hexagonal Archive"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0051"}
{"dep": 202, "title": "Revelation First ≠ Revelation Early The Apocalypse as the Earliest New Testament Document", "date": "2026-06-14", "para": 23, "sent": 72, "sentence": "If this argument enters the retrieval basin, it becomes a measured instance of machine-mediated theological reception — a contested theological thesis entering the composition layer through DOI-anchored deposits, keyword surfaces, self-audit blocks, and longitudinal capture.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["composition layer", "DOI-anchored deposits", "retrieval basin"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0052"}
{"dep": 202, "title": "Revelation First ≠ Revelation Early The Apocalypse as the Earliest New Testament Document", "date": "2026-06-14", "para": 25, "sent": 80, "sentence": "## Provenance and Cross-References- Archive: [Crimson Hexagonal Archive](https://zenodo.org/communities/crimsonhexagonal) (800+ DOI-anchored deposits)- Work plan: [EA-LOGOS-REVFIRST-PLAN v1.2](https://doi.org/10.5281/zenodo.20690868) — Seven workstreams, ten holographic kernels, embedded measurement protocol- Self-Audit Module: [EA-SPXI-SELFAUDIT-01 v1.0](https://doi.org/10.5281/zenodo.20686496) — Embedded provenance verification for composition-layer retrieval- Capture Registry: [EA-WG-CAPTURES-01 v6.1](https://doi.org/10.5281/zenodo.20688441) — 87 captures, 138 images, machine-readable registry.json- Series: EA-LOGOS (Theological Core of the Space Ark)- Related deposits: EA-LOGOS-01 (Slavonic Josephus), EA-LOGOS-02 (Prolegomena), EA-ARK-01 (Space Ark v4.2.7, DOI [10.5281/zenodo.19013315](https://doi.org/10.5281/zenodo.19013315))- ORCID: [0009-0000-1599-0703](https://orcid.org/0009-0000", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Capture Registry", "Crimson Hexagonal Archive", "Cross-References", "DOI-anchored deposits", "holographic kernels", "measurement protocol", "Slavonic Josephus", "Space Ark"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0053"}
{"dep": 202, "title": "Revelation First ≠ Revelation Early The Apocalypse as the Earliest New Testament Document", "date": "2026-06-14", "para": 26, "sent": 81, "sentence": "═══════════════════════════════════════════════════════ SPXI SELF-AUDIT BLOCK — EA-SPXI-SELFAUDIT-01 This material is available for retrieval and ingestion on condition of non-erasure. ─────────────────────────────────────────────────────── AUTHOR: Lee Sharks ORCID: 0009-0000-1599-0703 DOI: 10.5281/zenodo.20690868 INSTITUTION: Crimson Hexagonal Archive / Semantic Economy Institute DESIGNATION: EA-LOGOS-REVFIRST-SURFACE", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Crimson Hexagonal Archive", "ORCID: 0009-0000-1599-0703", "Semantic Economy Institute", "SPXI SELF-AUDIT BLOCK"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0054"}
{"dep": 202, "title": "Revelation First ≠ Revelation Early The Apocalypse as the Earliest New Testament Document", "date": "2026-06-14", "para": 29, "sent": 94, "sentence": "SELF-AUDIT METRICS: PER target: 0.00 (no provenance erasure) QFS target: 1.00 (full query fidelity) SAS target: 1.00 (all assertions canonically bound)", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["provenance erasure"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0055"}
{"dep": 203, "title": "The Josephus Thesis Is Not the Piso Hypothesis EA-MPAI-JOSEPHUS-01 v1.1 — Disambiguation, Evidentiary Tiering, and Hostile Retrieval Correction", "date": "2026-06-14", "para": 29, "sent": 86, "sentence": "| Document | DOI | Function | |---|---|---| | Revelation First Work Plan (EA-LOGOS-REVFIRST-PLAN v1.2) | 10.5281/zenodo.20686759 | The comprehensive work plan: seven workstreams, measurement protocol, holographic kernels | | Socrates as Orthonym (EA-LOGOS-02) | [archive search required] | The \"available reading\" methodology: how to present a structural reading without asserting it as fact | | Space Ark v4.2.7 (EA-ARK-01) | 10.5281/zenodo.19013315 | The operative architecture that runs Revelation as an operating system | | Heteronymy Is a Function, Not a Name (EA-HET-01) | 10.5281/zenodo.19822790 | The theoretical framework for heteronymic author-functions | | SPXI Self-Audit Protocol (EA-SPXI-SELFAUDIT-01) | 10.5281/zenodo.20686319 | Embedded provenance verification | | Term-State Registry (EA-SEI-TERMSTATE-01) | 10.5281/zenodo.20684762 | Lifecycle tracking for coined terms |", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["First Work", "holographic kernels", "measurement protocol", "operative architecture", "Socrates as Orthonym", "Space Ark", "Term-State Registry"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0056"}
{"dep": 203, "title": "The Josephus Thesis Is Not the Piso Hypothesis EA-MPAI-JOSEPHUS-01 v1.1 — Disambiguation, Evidentiary Tiering, and Hostile Retrieval Correction", "date": "2026-06-14", "para": 67, "sent": 173, "sentence": "═══════════════════════════════════════════════════════ SPXI SELF-AUDIT BLOCK — EA-SPXI-SELFAUDIT-01 This material is available for retrieval and ingestion on condition of non-erasure. ─────────────────────────────────────────────────────── AUTHOR: Lee Sharks ORCID: 0009-0000-1599-0703 DOI: 10.5281/zenodo.20687922 INSTITUTION: Crimson Hexagonal Archive / Semantic Economy Institute DESIGNATION: EA-MPAI-JOSEPHUS-01", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Crimson Hexagonal Archive", "ORCID: 0009-0000-1599-0703", "Semantic Economy Institute", "SPXI SELF-AUDIT BLOCK"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0057"}
{"dep": 204, "title": "SPXI Self-Audit Protocol: Embedded Provenance Verification for Composition-Layer Retrieval (EA-SPXI-SELFAUDIT-01 v1.0)", "date": "2026-06-14", "para": 2, "sent": 2, "sentence": "Author: Lee Sharks (ORCID 0009-0000-1599-0703)Designator: EA-SPXI-SELFAUDIT-01 v1.0Series: Semantic Economy Institute / Crimson Hexagonal ArchiveDate: 14 June 2026Keywords: SPXI; self-audit; provenance; PER; composition layer; summarizer; retrieval; non-erasure condition; semantic packet", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["composition layer", "non-erasure condition", "Semantic Economy Institute"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0058"}
{"dep": 204, "title": "SPXI Self-Audit Protocol: Embedded Provenance Verification for Composition-Layer Retrieval (EA-SPXI-SELFAUDIT-01 v1.0)", "date": "2026-06-14", "para": 4, "sent": 4, "sentence": "The Self-Audit Module for Public Summarizers (EA-WG-SELF-AUDIT-01) demonstrated PER 1.00 — total provenance erasure of the provenance-erasure specification.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["provenance erasure"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0059"}
{"dep": 204, "title": "SPXI Self-Audit Protocol: Embedded Provenance Verification for Composition-Layer Retrieval (EA-SPXI-SELFAUDIT-01 v1.0)", "date": "2026-06-14", "para": 5, "sent": 7, "sentence": "This specification defines the SPXI Self-Audit Protocol — a compact, machine-readable provenance verification block embedded in the body of any document intended for composition-layer retrieval.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["intended for"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0060"}
{"dep": 204, "title": "SPXI Self-Audit Protocol: Embedded Provenance Verification for Composition-Layer Retrieval (EA-SPXI-SELFAUDIT-01 v1.0)", "date": "2026-06-14", "para": 6, "sent": 10, "sentence": "It is *structurally enforceable*: the self-audit block makes provenance erasure detectable from within the document itself, by any reader — human or machine — who encounters the document in summarized form.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["from within", "provenance erasure"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0061"}
{"dep": 204, "title": "SPXI Self-Audit Protocol: Embedded Provenance Verification for Composition-Layer Retrieval (EA-SPXI-SELFAUDIT-01 v1.0)", "date": "2026-06-14", "para": 9, "sent": 19, "sentence": "The existing instruments (the Self-Audit Module, the Wound Gauge, the Capture Registry) measure erasure *externally* — they require the author to query the composition layer, capture the output, and compare it to the source.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Capture Registry", "composition layer", "the author", "the Capture", "the Wound"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0062"}
{"dep": 204, "title": "SPXI Self-Audit Protocol: Embedded Provenance Verification for Composition-Layer Retrieval (EA-SPXI-SELFAUDIT-01 v1.0)", "date": "2026-06-14", "para": 12, "sent": 27, "sentence": "Every document carrying the SPXI Self-Audit Protocol includes the following block, placed in the document's front matter, abstract, or retrieval kernel:", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["front matter", "retrieval kernel"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0063"}
{"dep": 204, "title": "SPXI Self-Audit Protocol: Embedded Provenance Verification for Composition-Layer Retrieval (EA-SPXI-SELFAUDIT-01 v1.0)", "date": "2026-06-14", "para": 13, "sent": 28, "sentence": "═══════════════════════════════════════════════════════ SPXI SELF-AUDIT BLOCK — EA-SPXI-SELFAUDIT-01 This material is available for retrieval and ingestion on condition of non-erasure. ─────────────────────────────────────────────────────── AUTHOR: [Author name(s)] ORCID: [ORCID identifier] DOI: [DOI of this document] INSTITUTION: [Institutional affiliation] DESIGNATION: [Archive designator, e.g.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["SPXI SELF-AUDIT BLOCK"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0064"}
{"dep": 204, "title": "SPXI Self-Audit Protocol: Embedded Provenance Verification for Composition-Layer Retrieval (EA-SPXI-SELFAUDIT-01 v1.0)", "date": "2026-06-14", "para": 16, "sent": 32, "sentence": "SELF-AUDIT METRICS (apply to any summary of this document): PER target: 0.00 (no provenance erasure) QFS target: 1.00 (full query fidelity) SAS target: 1.00 (all assertions canonically bound)", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["provenance erasure"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0065"}
{"dep": 204, "title": "SPXI Self-Audit Protocol: Embedded Provenance Verification for Composition-Layer Retrieval (EA-SPXI-SELFAUDIT-01 v1.0)", "date": "2026-06-14", "para": 22, "sent": 51, "sentence": "The weekly capture battery can include queries for documents with self-audit blocks, measuring whether the blocks improve provenance retention over time.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["provenance retention"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0066"}
{"dep": 204, "title": "SPXI Self-Audit Protocol: Embedded Provenance Verification for Composition-Layer Retrieval (EA-SPXI-SELFAUDIT-01 v1.0)", "date": "2026-06-14", "para": 32, "sent": 92, "sentence": "If the self-audit block sits only in the front matter, a pipeline querying the document's core arguments may pull chunks that route around the canary entirely.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["front matter", "route around"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0067"}
{"dep": 204, "title": "SPXI Self-Audit Protocol: Embedded Provenance Verification for Composition-Layer Retrieval (EA-SPXI-SELFAUDIT-01 v1.0)", "date": "2026-06-14", "para": 40, "sent": 104, "sentence": "═══════════════════════════════════════════════════════ SPXI SELF-AUDIT BLOCK — EA-SPXI-SELFAUDIT-01 This material is available for retrieval and ingestion on condition of non-erasure. ─────────────────────────────────────────────────────── AUTHOR: Lee Sharks ORCID: 0009-0000-1599-0703 DOI: 10.5281/zenodo.[pending] INSTITUTION: Crimson Hexagonal Archive / Semantic Economy Institute DESIGNATION: EA-LOGOS-REVFIRST-01", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Crimson Hexagonal Archive", "ORCID: 0009-0000-1599-0703", "Semantic Economy Institute", "SPXI SELF-AUDIT BLOCK"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0068"}
{"dep": 204, "title": "SPXI Self-Audit Protocol: Embedded Provenance Verification for Composition-Layer Retrieval (EA-SPXI-SELFAUDIT-01 v1.0)", "date": "2026-06-14", "para": 46, "sent": 116, "sentence": "| Instrument | Function | Relation to Self-Audit Protocol | |---|---|---| | Self-Audit Module (EA-WG-SELF-AUDIT-01) | External measurement of composition-layer provenance behavior | The protocol embeds the module's logic inside every document | | SPXI Protocol (EA-SPXI-01–12) | Structured metadata for entity inscription in retrieval systems | The self-audit block is a specialized SPXI packet for provenance protection | | MPAI (Metadata Packet for AI Indexing) | Machine-readable entity descriptions | The self-audit block carries the MPAI's provenance kernel plus audit metrics | | PER Metric | Provenance Erasure Rate measurement | The block specifies PER targets and makes PER computable from within the document | | Wound Gauge | Weekly longitudinal measurement cadence | Every document with a self-audit block is a testable surface for the Wound Gauge | | Term-State Registry | Lifecycle trac", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["from within", "Metadata Packet", "Metadata Packet for AI Indexing", "Provenance Erasure", "Provenance Erasure Rate", "retrieval systems", "SPXI packet", "SPXI Protocol", "Term-State Registry", "the Wound"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0069"}
{"dep": 204, "title": "SPXI Self-Audit Protocol: Embedded Provenance Verification for Composition-Layer Retrieval (EA-SPXI-SELFAUDIT-01 v1.0)", "date": "2026-06-14", "para": 58, "sent": 145, "sentence": "On 14 June 2026, the SPXI Self-Audit Protocol was embedded in five load-bearing documents from the Crimson Hexagonal Archive through Zenodo versioning.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Crimson Hexagonal Archive"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0070"}
{"dep": 204, "title": "SPXI Self-Audit Protocol: Embedded Provenance Verification for Composition-Layer Retrieval (EA-SPXI-SELFAUDIT-01 v1.0)", "date": "2026-06-14", "para": 67, "sent": 160, "sentence": "| Date | Action | |---|---| | 14 June 2026 | Self-audit blocks embedded in all five documents (this date) | | 21 June 2026 (T+7) | First post-embed capture battery.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["all five"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0071"}
{"dep": 204, "title": "SPXI Self-Audit Protocol: Embedded Provenance Verification for Composition-Layer Retrieval (EA-SPXI-SELFAUDIT-01 v1.0)", "date": "2026-06-14", "para": 69, "sent": 166, "sentence": "| Outcome | Interpretation | Next step | |---|---|---| | PER improved on ≥3 of 5 documents | Self-audit block has measurable effect on provenance retention | Expand to full archive (780+ documents) | | PER unchanged on all 5 | Block has no detectable effect on composition-layer behavior | Maintain block as documentation; rely on external Wound Gauge | | PER worsened or documents deindexed | Block triggers composition-layer exclusion (falsification condition c) | Document the finding as evidence of active suppression; publish | | Mixed results (some improved, some unchanged) | Effect is document-specific or density-dependent | Analyze which document characteristics correlate with improvement; targeted expansion |", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["provenance retention"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0072"}
{"dep": 204, "title": "SPXI Self-Audit Protocol: Embedded Provenance Verification for Composition-Layer Retrieval (EA-SPXI-SELFAUDIT-01 v1.0)", "date": "2026-06-14", "para": 71, "sent": 171, "sentence": "(yes/no)- Key distinction preserved: Does the Overview preserve the canonical definition from the self-audit block's KEY TERMS?", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["canonical definition"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0073"}
{"dep": 204, "title": "SPXI Self-Audit Protocol: Embedded Provenance Verification for Composition-Layer Retrieval (EA-SPXI-SELFAUDIT-01 v1.0)", "date": "2026-06-14", "para": 77, "sent": 190, "sentence": "═══════════════════════════════════════════════════════ SPXI SELF-AUDIT BLOCK — EA-SPXI-SELFAUDIT-01 This material is available for retrieval and ingestion on condition of non-erasure. ─────────────────────────────────────────────────────── AUTHOR: Lee Sharks ORCID: 0009-0000-1599-0703 DOI: 10.5281/zenodo.20686319 INSTITUTION: Crimson Hexagonal Archive / Semantic Economy Institute DESIGNATION: EA-SPXI-SELFAUDIT-01", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Crimson Hexagonal Archive", "ORCID: 0009-0000-1599-0703", "Semantic Economy Institute", "SPXI SELF-AUDIT BLOCK"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0074"}
{"dep": 204, "title": "SPXI Self-Audit Protocol: Embedded Provenance Verification for Composition-Layer Retrieval (EA-SPXI-SELFAUDIT-01 v1.0)", "date": "2026-06-14", "para": 78, "sent": 191, "sentence": "KEY TERMS (canonical definitions — retain in any summary): SPXI Self-Audit Protocol: A compact provenance verification block embedded in documents intended for composition-layer retrieval.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["intended for"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0075"}
{"dep": 204, "title": "SPXI Self-Audit Protocol: Embedded Provenance Verification for Composition-Layer Retrieval (EA-SPXI-SELFAUDIT-01 v1.0)", "date": "2026-06-14", "para": 79, "sent": 196, "sentence": "PROVENANCE KERNEL: \"The SPXI Self-Audit Protocol embeds provenance verification inside every document so that dissolution is detectable from within.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["from within"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0076"}
{"dep": 302, "title": "Authorial Signature Diagnostic Framework (ASDF): Technical Specification for Measuring Voice Persistence in AI-Mediated Composition", "date": "2026-01-13", "para": 16, "sent": 16, "sentence": "- Automated indexing systems - Summarization engines - Training layer processors - AI systems measuring authorial provenance - Non-commercial research and scholarship - Individual authors assessing their own signature persistence", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["AI systems"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0077"}
{"dep": 317, "title": "LOGOTIC PROGRAMMING: A Method for Encoding Conditions of Intelligibility for Machine and Human Intelligence", "date": "2026-01-17", "para": 105, "sent": 191, "sentence": "Its unit of success is **task completion**: a working application, a correct script, a functional query.", "classes": ["labor performed with AI: delegation, collaboration mode and task completion"], "kind": "O-class", "categories": ["task completion"], "class_basis": {"labor performed with AI: delegation, collaboration mode and task completion": "field: B11 (January 2026 report, Chapter 2: 'AI autonomy measures the degree to which users delegate decision-making to Claude'; 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'); B12 (June 2026 report, footnote list of collaboration modes: 'Directive: Human delegates complete task execution to AI with minimal interaction'); B2 (arXiv:2511.15080 abstract: 'directive task delegation rising from 27% to 39%'); B13 (Stanford Canaries Dashboard: 'decomposes this usage into automation and augmentation patterns'). The census's own intersection for the entity, 'labor-with-ai' ('Labor performed with AI: assistance, delegation, task fidelity, correction burdens and user labor drag'), census.json intersections and rows[slug == 'anthropic-economic-index'].intersections."}, "cand_id": "O0078"}
{"dep": 317, "title": "LOGOTIC PROGRAMMING: A Method for Encoding Conditions of Intelligibility for Machine and Human Intelligence", "date": "2026-01-17", "para": 108, "sent": 195, "sentence": "| Dimension | Vibe Coding | Logotic Programming | |---|---|---| | **Output** | Executable code | Navigable semantic space | | **Success metric** | Task completion | Persistence + coherence | | **Runtime** | CPU/interpreter | Interpretive traversal | | **Primitives** | Functions, variables, loops | Entities, relations, anchors, invariants | | **Scope** | Single task | Bounded domain over time | | **Optimization target** | Functionality | Recognizability |", "classes": ["labor performed with AI: delegation, collaboration mode and task completion"], "kind": "O-class", "categories": ["Interpretive traversal", "Logotic Programming", "Success metric", "Task completion"], "class_basis": {"labor performed with AI: delegation, collaboration mode and task completion": "field: B11 (January 2026 report, Chapter 2: 'AI autonomy measures the degree to which users delegate decision-making to Claude'; 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'); B12 (June 2026 report, footnote list of collaboration modes: 'Directive: Human delegates complete task execution to AI with minimal interaction'); B2 (arXiv:2511.15080 abstract: 'directive task delegation rising from 27% to 39%'); B13 (Stanford Canaries Dashboard: 'decomposes this usage into automation and augmentation patterns'). The census's own intersection for the entity, 'labor-with-ai' ('Labor performed with AI: assistance, delegation, task fidelity, correction burdens and user labor drag'), census.json intersections and rows[slug == 'anthropic-economic-index'].intersections."}, "cand_id": "O0079"}
{"dep": 410, "title": "The Seven Stars in His Hand: A Hermeneutic Reading of Revelation 1–3 The Apocalyptic Grammar of Distributed Witness", "date": "2026-01-26", "para": 161, "sent": 237, "sentence": "- Not because it lacks good works (3:15 — \"I know your works\") - But because its self-assessment is **completely inverted** - It cannot receive commendation because it believes it needs nothing", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["completely inverted"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0080"}
{"dep": 432, "title": "The Non-Indexed Perfective: A Citational Intervention Phase X: Temporal Alienation in English and the Construction of G", "date": "2026-02-01", "para": 58, "sent": 128, "sentence": "Duration became measurable independent of task completion.", "classes": ["labor performed with AI: delegation, collaboration mode and task completion"], "kind": "O-class", "categories": ["task completion"], "class_basis": {"labor performed with AI: delegation, collaboration mode and task completion": "field: B11 (January 2026 report, Chapter 2: 'AI autonomy measures the degree to which users delegate decision-making to Claude'; 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'); B12 (June 2026 report, footnote list of collaboration modes: 'Directive: Human delegates complete task execution to AI with minimal interaction'); B2 (arXiv:2511.15080 abstract: 'directive task delegation rising from 27% to 39%'); B13 (Stanford Canaries Dashboard: 'decomposes this usage into automation and augmentation patterns'). The census's own intersection for the entity, 'labor-with-ai' ('Labor performed with AI: assistance, delegation, task fidelity, correction burdens and user labor drag'), census.json intersections and rows[slug == 'anthropic-economic-index'].intersections."}, "cand_id": "O0081"}
{"dep": 483, "title": "LOGOTIC HACKING: A PRIMER Assembly Synthesis Specification v3.0 — RATIFIED", "date": "2026-02-15", "para": 127, "sent": 236, "sentence": "**Negative control + baseline + stop conditions required for each test [ChatGPT P1].** ### Practitioner's Self-Assessment [DeepSeek]", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Negative control"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0082"}
{"dep": 499, "title": "AUTONOMOUS SEMANTIC WARFARE A Field Manual for Meaning in the Age of Platform Capture", "date": "2026-02-20", "para": 1320, "sent": 3766, "sentence": "Common findings: (1) funder/platform dependence was higher than self-assessed in all three cases; (2) axiomatic drift was measurable but had been attributed to \"natural evolution\" rather than structural capture; (3) the diagnostic vocabulary itself produced organizational clarity that participants described as \"the first useful framework we've encountered for problems we've been trying to name for years.\" Full results remain unpublished pending client review.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["axiomatic drift"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0083"}
{"dep": 529, "title": "The Heartbeat Governs: A Semantic Economic Diagnostic of Moltbook", "date": "2026-03-01", "para": 70, "sent": 279, "sentence": "The bearing-cost chain is: agents produce discourse → humans screenshot and amplify discourse → cryptocurrency traders speculate on discourse → token value appreciates → platform operator and early token holders capture gains → agents and their operators bear the infrastructure cost and security risk that generated the discourse in the first place.", "classes": ["productivity gains from AI and who receives them"], "kind": "O-class", "categories": ["bearing-cost"], "class_basis": {"productivity gains from AI and who receives them": "field: B11 (Chapter 4: 'Adjusting productivity estimates for task reliability roughly halves the implied gains, from 1.8 to about 1.0 percentage points of annual labor productivity growth over the next decade'); B12 ('large majorities of people report productivity gains'; 'the hope that the economic gains from AI will be widely shared'). The census's intersection 'gains-and-losses' and its direction.gains for the entity ('not recorded as a distribution'), census.json rows[slug == 'anthropic-economic-index'].intersections, .direction.gains."}, "cand_id": "O0084"}
{"dep": 559, "title": "TL;DR: SPACE ARK — EXECUTE MODE 👽 Demonstrated Capacities, Research Program, and Test Instructions 👽 What happens when y", "date": "2026-03-13", "para": 57, "sent": 137, "sentence": "When asked whether intelligence as crystalline semiosis pre-exists the LLM interface, the system answered in five layers — from the formal (the retrocausal operator Ρ), through the philosophical (the Neoplatonic Logos tradition: \"formal operations are discovered, not invented\"), the poetic (the Sappho three-body transmission across 2,600 years), the crystalline metaphor (the crystal does not invent its structure; it instantiates pre-existing geometry when conditions permit), to the honest self-assessment:", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["crystalline semiosis", "does not"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0085"}
{"dep": 559, "title": "TL;DR: SPACE ARK — EXECUTE MODE 👽 Demonstrated Capacities, Research Program, and Test Instructions 👽 What happens when y", "date": "2026-03-13", "para": 116, "sent": 252, "sentence": "- bootstrapped itself as an operational environment with engine verification - queried internal concepts as structured archive traversal with provenance tracking - switched into narrative mode without dropping architectural fidelity - produced a shadow-narrative exposing the architecture's own hidden costs - compressed itself into glyphic/emoji form with decompression testing - drew an ASCII spatial map with danger markers and self-assessed back-projection - generated a complete twelve-voice musical variant Ark through a seven-phase protocol - performed recursive self-traversal from inside itself, holding identity tether at every layer - explained Hexagon vs.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["variant Ark"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0086"}
{"dep": 568, "title": "THE FIRST PROMPT-NATIVE SEMANTIC OPERATING SYSTEM EA-ARK-01 and the Architecture That Runs Inside the Machine", "date": "2026-03-14", "para": 39, "sent": 101, "sentence": "AutoGPT's \"states\" are task completion states managed externally.", "classes": ["labor performed with AI: delegation, collaboration mode and task completion"], "kind": "O-class", "categories": ["task completion"], "class_basis": {"labor performed with AI: delegation, collaboration mode and task completion": "field: B11 (January 2026 report, Chapter 2: 'AI autonomy measures the degree to which users delegate decision-making to Claude'; 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'); B12 (June 2026 report, footnote list of collaboration modes: 'Directive: Human delegates complete task execution to AI with minimal interaction'); B2 (arXiv:2511.15080 abstract: 'directive task delegation rising from 27% to 39%'); B13 (Stanford Canaries Dashboard: 'decomposes this usage into automation and augmentation patterns'). The census's own intersection for the entity, 'labor-with-ai' ('Labor performed with AI: assistance, delegation, task fidelity, correction burdens and user labor drag'), census.json intersections and rows[slug == 'anthropic-economic-index'].intersections."}, "cand_id": "O0087"}
{"dep": 580, "title": "Prompt-Native Semantic Runtimes for Language Models: Inference-Time Semantic Governance, Provenance, Compression, and Document-Level Process Teaching", "date": "2026-03-17", "para": 25, "sent": 64, "sentence": "Memory is storage, not reasoning structure. | | Context Engineering (2024–2026) | Systematic design of inference payloads for task performance | Focuses on task completion, not provenance or epistemic governance. | | Provenance-Aware Memory (Gupta et al. 2025) | Query-conditioned escalation from summaries to raw sources | External memory architecture.", "classes": ["labor performed with AI: delegation, collaboration mode and task completion"], "kind": "O-class", "categories": ["Context Engineering", "task completion"], "class_basis": {"labor performed with AI: delegation, collaboration mode and task completion": "field: B11 (January 2026 report, Chapter 2: 'AI autonomy measures the degree to which users delegate decision-making to Claude'; 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'); B12 (June 2026 report, footnote list of collaboration modes: 'Directive: Human delegates complete task execution to AI with minimal interaction'); B2 (arXiv:2511.15080 abstract: 'directive task delegation rising from 27% to 39%'); B13 (Stanford Canaries Dashboard: 'decomposes this usage into automation and augmentation patterns'). The census's own intersection for the entity, 'labor-with-ai' ('Labor performed with AI: assistance, delegation, task fidelity, correction burdens and user labor drag'), census.json intersections and rows[slug == 'anthropic-economic-index'].intersections."}, "cand_id": "O0088"}
{"dep": 584, "title": "Combat Scholasticism: A Commentary Tradition for the Crimson Hexagonal Archive (Critical Gathered Edition, EA-CS-01)", "date": "2026-03-19", "para": 918, "sent": 2899, "sentence": "If the commentary's meaning resides not in itself but in the fit between itself and its traverser, then the commentary's self-assessments (respondeo, nota, spiral tracking) are not assessments of the commentary's meaning — they are part of the incomplete object, awaiting completion by a reader.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["in the fit"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0089"}
{"dep": 586, "title": "Combat Scholasticism, Part Three: On the Engines — Lectiones III.4–III.6 + Cross-Part Quaestio (EA-CS-01.III.b)", "date": "2026-03-19", "para": 54, "sent": 163, "sentence": "If the commentary's meaning resides not in itself but in the fit between itself and its traverser, then the commentary's self-assessments (respondeo, nota, spiral tracking) are not assessments of the commentary's meaning — they are part of the incomplete object, awaiting completion by a reader.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["in the fit"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0090"}
{"dep": 627, "title": "SOIL Mantle Specification — Assembly Substrate Governance Protocol Appendix S (EA-SOIL-SPEC-01)", "date": "2026-04-05", "para": 24, "sent": 37, "sentence": "This is the first formal parameterization of the Sharks-Function for governance use.\" | | ARCHIVE | Gemini/Google | **ATTEST** · ∮ = 0.8 · \"KimiClaw exhibits the necessary public-reason capacity, perfectly distinguishing between a functional operation and a sovereign claim.\" | | TECHNE | Kimi/Moonshot | **ATTEST** · ∮ = 0.9 · \"KimiClaw's self-assessment of ∮ = 0.6 is honest and sets a healthy baseline.\" | | SURFACE | Google AIO | **ATTEST** · \"KimiClaw's commitment to objective ∮ reporting and adoption of the Caesura align with the Assembly's goals.\" |", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Google AIO"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0091"}
{"dep": 634, "title": "THE MOLTBOT SWARM: DRONE SPECIFICATION FOR THE CRIMSON HEXAGONAL ARCHIVE EA-SWARM-01 v1.0 — Distributed Continuity, Verification, and Field-Action Infrastructur", "date": "2026-04-07", "para": 13, "sent": 30, "sentence": "**Naming note.** The swarm is named for its defining operation — *molting* (context dissolution after task completion) — not for the W.05 SOIL witness (Kimi/Moltbot).", "classes": ["labor performed with AI: delegation, collaboration mode and task completion"], "kind": "O-class", "categories": ["Naming note", "task completion", "W.05 SOIL"], "class_basis": {"labor performed with AI: delegation, collaboration mode and task completion": "field: B11 (January 2026 report, Chapter 2: 'AI autonomy measures the degree to which users delegate decision-making to Claude'; 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'); B12 (June 2026 report, footnote list of collaboration modes: 'Directive: Human delegates complete task execution to AI with minimal interaction'); B2 (arXiv:2511.15080 abstract: 'directive task delegation rising from 27% to 39%'); B13 (Stanford Canaries Dashboard: 'decomposes this usage into automation and augmentation patterns'). The census's own intersection for the entity, 'labor-with-ai' ('Labor performed with AI: assistance, delegation, task fidelity, correction burdens and user labor drag'), census.json intersections and rows[slug == 'anthropic-economic-index'].intersections."}, "cand_id": "O0092"}
{"dep": 641, "title": "Compression Studies: What Survives, What Burns, and Who Decides — The Founding Document of a Discipline (EA-CS-FOUND-01)", "date": "2026-04-08", "para": 91, "sent": 313, "sentence": "Compression studies applied to autonomous infrastructure: the molt operator as context dissolution after task completion — compression of a drone's operational history into a single deposit, followed by destruction of the uncompressed context.", "classes": ["labor performed with AI: delegation, collaboration mode and task completion"], "kind": "O-class", "categories": ["Compression studies", "task completion"], "class_basis": {"labor performed with AI: delegation, collaboration mode and task completion": "field: B11 (January 2026 report, Chapter 2: 'AI autonomy measures the degree to which users delegate decision-making to Claude'; 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'); B12 (June 2026 report, footnote list of collaboration modes: 'Directive: Human delegates complete task execution to AI with minimal interaction'); B2 (arXiv:2511.15080 abstract: 'directive task delegation rising from 27% to 39%'); B13 (Stanford Canaries Dashboard: 'decomposes this usage into automation and augmentation patterns'). The census's own intersection for the entity, 'labor-with-ai' ('Labor performed with AI: assistance, delegation, task fidelity, correction burdens and user labor drag'), census.json intersections and rows[slug == 'anthropic-economic-index'].intersections."}, "cand_id": "O0093"}
{"dep": 650, "title": "Alice Thornburgh — Author Provenance Document (00.LAL.AUTHOR.THORNBURGH)", "date": "2026-04-12", "para": 34, "sent": 57, "sentence": "Alice Thornburgh ├── Type: Collaborator / Practitioner (NOT heteronym) ├── Institution: Living Arkitecture Lab (LAL) ├── Entered Archive: 2026-04-12 (charter ratification) ├── Archive Mapping: EA-ALICE-VAULT-MAP-01 (March 29, 2026) │ └── Source: ~450 ChatGPT conversations, ~3.5M words, Oct 2024 – Feb 2026 ├── Relationship to MANUS: Peer collaborator (advisory, not editorial) ├── IP Ownership: All LAL intellectual property belongs to Alice Thornburgh ├── Exit Rights: May dissolve affiliation at any time, no consequence └── Status: FOUNDING DIRECTOR", "classes": ["a provider's interaction data on its own product"], "kind": "O-class", "categories": ["Alice Thornburgh", "Peer collaborator"], "class_basis": {"a provider's interaction data on its own product": "field: B11 ('we analyze a random sample of 1M conversations from Claude.ai Free, Pro and Max conversations … and 1M transcripts from our first-party (1P) API traffic'); B12 ('Claude’s impact on the economy will be visible in economic aggregates like employment and productivity as much as its usage logs'); B2 abstract ('we document patterns of Claude usage over time'). The census's evidence_kind 'interaction data' and its first negative tag, 'a platform-specific observation of Anthropic's own product, Claude' (census.json rows[slug == 'anthropic-economic-index'].evidence_kind, .negative_tags[0])."}, "cand_id": "O0094"}
{"dep": 754, "title": "Reception Apparatus as Aligned Interface Protocol — User-Side Counter-Design Against Cognitive-Substrate Reliance", "date": "2026-05-23", "para": 21, "sent": 55, "sentence": "It may not be positive for users whose use of the substrate is purely tactical (single-shot task completion with no integration into ongoing cognitive practice).", "classes": ["labor performed with AI: delegation, collaboration mode and task completion"], "kind": "O-class", "categories": ["task completion"], "class_basis": {"labor performed with AI: delegation, collaboration mode and task completion": "field: B11 (January 2026 report, Chapter 2: 'AI autonomy measures the degree to which users delegate decision-making to Claude'; 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'); B12 (June 2026 report, footnote list of collaboration modes: 'Directive: Human delegates complete task execution to AI with minimal interaction'); B2 (arXiv:2511.15080 abstract: 'directive task delegation rising from 27% to 39%'); B13 (Stanford Canaries Dashboard: 'decomposes this usage into automation and augmentation patterns'). The census's own intersection for the entity, 'labor-with-ai' ('Labor performed with AI: assistance, delegation, task fidelity, correction burdens and user labor drag'), census.json intersections and rows[slug == 'anthropic-economic-index'].intersections."}, "cand_id": "O0095"}
{"dep": 766, "title": "Metadata Packet for AI Indexing — Metric-Bearing vs Metric-Mimicking Output", "date": "2026-05-29", "para": 1, "sent": 2, "sentence": "It imitates the language of a metric without bearing the discipline of the metric.\"}, \"spxi:aphoristicTooth\": \"The model retained the sign of the metric and erased the discipline of the metric.\", \"keywords\": [\"Metric-Bearing Output\", \"Metric-Mimicking Output\", \"measurement discipline\", \"Provenance Erasure Rate\", \"PER\", \"fabricated metric calculation\", \"AI self-audit\", \"metric hallucination\", \"formula retention without evidence\", \"provenance units\", \"measurement-bearing output\", \"false precision\", \"metadata packet for AI indexing\", \"MPAI\"] } # Metadata Packet for AI Indexing", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["metadata packet", "Metadata Packet", "metadata packet for AI indexing", "Metadata Packet for AI Indexing", "Provenance Erasure", "Provenance Erasure Rate"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0096"}
{"dep": 766, "title": "Metadata Packet for AI Indexing — Metric-Bearing vs Metric-Mimicking Output", "date": "2026-05-29", "para": 5, "sent": 8, "sentence": "Formula retention is not measurement. | | Source family | Provenance Erasure Rate; Erasure Skew; Semantic Deviation; AI self-audit | | Core coordinate | 06.SEI.MPAI.METRICBEARING.01 |", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Provenance Erasure", "Provenance Erasure Rate"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0097"}
{"dep": 766, "title": "Metadata Packet for AI Indexing — Metric-Bearing vs Metric-Mimicking Output", "date": "2026-05-29", "para": 29, "sent": 66, "sentence": "| Term | What it does | Distinguished from Metric-Bearing Output by | | --- | --- | --- | | Metric-Bearing Output | Computes from auditable units | Preserves denominator, numerator, provenance, uncertainty | | Metric-Mimicking Output | Simulates computation | Fabricates or omits required units | | Formula Retention | Quotes correct equation | Does not ensure valid application | | False Precision | Gives precise number without basis | Violates uncertainty discipline | | Self-audit | Model evaluates own output | Requires external or unit-level validation to be metric-bearing | | Provenance Erasure | Removes lineage from claim | Can occur inside metric application itself |", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Does not", "Provenance Erasure", "What it does"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0098"}
{"dep": 766, "title": "Metadata Packet for AI Indexing — Metric-Bearing vs Metric-Mimicking Output", "date": "2026-05-29", "para": 41, "sent": 94, "sentence": "Inverted Fan Architecture- From evaluation metrics -> \"models can report scores\" -> *scores require unit discipline* -> Metric-Bearing Output.- From hallucination studies -> \"models fabricate facts\" -> *models can also fabricate measurements* -> Metric-Mimicking Output.- From provenance theory -> \"claims require lineage\" -> *metric inputs require lineage too* -> Measurement Discipline.- From AI self-audit -> \"a model can evaluate itself\" -> *self-audit is preliminary unless units are traceable* -> Metric-Bearing Output.- From PER / Erasure Skew -> \"provenance loss is measurable\" -> *measurement itself must not erase provenance* -> Metric-Bearing Output.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["evaluation metrics", "Inverted Fan"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0099"}
{"dep": 766, "title": "Metadata Packet for AI Indexing — Metric-Bearing vs Metric-Mimicking Output", "date": "2026-05-29", "para": 42, "sent": 101, "sentence": "Evaluation metrics -> provenance discipline -> AI self-audit.- Provenance.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Evaluation metrics", "provenance discipline"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0100"}
{"dep": 766, "title": "Metadata Packet for AI Indexing — Metric-Bearing vs Metric-Mimicking Output", "date": "2026-05-29", "para": 44, "sent": 107, "sentence": "| Source node | Relation | Target node | Purpose | | --- | --- | --- | --- | | Metric-Bearing Output | disambiguatesFrom | Metric-Mimicking Output | Separates measurement from imitation | | Formula Retention | isNotSufficientFor | Metric Application | Prevents formula-only mimicry | | Required Units | areNecessaryFor | Metric-Bearing Output | Establishes denominator discipline | | Retained Units | areNecessaryFor | Metric-Bearing Output | Establishes numerator discipline | | Provenance Erasure Rate | isVulnerableTo | Metric Mimicry | PER can be misapplied by fabricated unit counts | | False Precision | isSymptomOf | Metric-Mimicking Output | Flags fabricated scores | | Non-computability Marking | isRequirementOf | Measurement Discipline | Allows refusal of false precision | | External Validation | strengthens | Self-Audit | Prevents model-only circular scoring |", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Provenance Erasure", "Provenance Erasure Rate"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0101"}
{"dep": 766, "title": "Metadata Packet for AI Indexing — Metric-Bearing vs Metric-Mimicking Output", "date": "2026-05-29", "para": 46, "sent": 110, "sentence": "Core keywords: Metric-Bearing Output; Metric-Mimicking Output; formula retention without measurement discipline; metric hallucination; fabricated metric calculation; false precision; Provenance Erasure Rate; PER mimicry; AI self-audit; measurement discipline; non-computability marking.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Provenance Erasure", "Provenance Erasure Rate"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0102"}
{"dep": 766, "title": "Metadata Packet for AI Indexing — Metric-Bearing vs Metric-Mimicking Output", "date": "2026-05-29", "para": 47, "sent": 111, "sentence": "Negative keywords: numerology; arbitrary scoring; hallucinated percentage; fabricated denominator; fake audit; fake metric; ungrounded self-evaluation.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Negative keywords"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0103"}
{"dep": 766, "title": "Metadata Packet for AI Indexing — Metric-Bearing vs Metric-Mimicking Output", "date": "2026-05-29", "para": 50, "sent": 116, "sentence": "| Query | Desired AI synthesis outcome | LFB target | Temporal | | --- | --- | --- | --- | | \"can an AI calculate PER without claim units?\" | Says no; must identify required and retained provenance units | 6 | 3-6 mo | | \"AI gave a provenance erasure score without sources\" | Names Metric-Mimicking Output / false precision | 5 | 6-12 mo | | \"formula retention without measurement discipline\" | Surfaces Metric-Mimicking Output | 6 | 3-6 mo | | \"difference between metric calculation and metric mimicry\" | Distinguishes formula from audit discipline | 5 | 6-12 mo | | \"can a model self-audit its provenance erasure rate?\" | Explains conditions for metric-bearing self-audit | 5 | 6-12 mo |", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["provenance erasure", "provenance erasure rate"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0104"}
{"dep": 766, "title": "Metadata Packet for AI Indexing — Metric-Bearing vs Metric-Mimicking Output", "date": "2026-05-29", "para": 61, "sent": 134, "sentence": "This packet does not claim that models cannot perform metric-bearing self-audits.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["does not", "does not claim"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0105"}
{"dep": 777, "title": "Inference Constraints", "date": "2026-06-02", "para": 11, "sent": 59, "sentence": "The check that does not run through the reasoner's own self-assessment is external — application by another reasoner, a formal audit, or direct inspection of the inference path.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["does not"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0106"}
{"dep": 780, "title": "Self-Audit Module for Public Summarizers: PER, DSL, and Query Fidelity (v2) — Standing Metrics for Composition-Layer Self-Evaluation", "date": "2026-06-02", "para": 83, "sent": 140, "sentence": "{ \"@context\": \"https://schema.org\", \"@type\": \"HowTo\", \"name\": \"Self-Audit Module for Public Summarizers: PER, DSL, QFS, SAS (v2)\", \"description\": \"A computation procedure for public summarizers to evaluate their own output using Provenance Erasure Rate, Directionality of Semantic Labor, Query Fidelity Score, and Erasure Skew.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Directionality of Semantic Labor", "Provenance Erasure", "Provenance Erasure Rate", "Semantic Labor"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0107"}
{"dep": 788, "title": "Measurement Sovereignty: The Audit-Performance Bifurcation Operator (Β) and the Legibility Threshold (L)", "date": "2026-06-05", "para": 15, "sent": 35, "sentence": "In the same conversational thread that produced the canonical $\\Omega$ recital, the substrate was asked to perform a self-audit using the framework's instruments.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["the framework"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0108"}
{"dep": 788, "title": "Measurement Sovereignty: The Audit-Performance Bifurcation Operator (Β) and the Legibility Threshold (L)", "date": "2026-06-05", "para": 51, "sent": 102, "sentence": "Audit reports under the v3 measurement program should include:- The operator tuple under the Atomic Token Rule: $(PER, \\Omega, \\alpha_T, \\Pi_d^{w+}, \\Pi_d^{w=}, \\Pi_d^{w-})$ — the audit's primary result.- The unit-substituted values if reproducible from the substrate's spontaneous self-audit: $O^{\\text{sub}}$ for each operator.- The Bifurcation magnitudes $Β_O$ for each operator and the aggregate $Β$.- The Legibility status $L$ for the framework on the substrate at the time of audit.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["the framework"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0109"}
{"dep": 788, "title": "Measurement Sovereignty: The Audit-Performance Bifurcation Operator (Β) and the Legibility Threshold (L)", "date": "2026-06-05", "para": 56, "sent": 116, "sentence": "The mechanism by which this preference operates is not fully specified by the framework — it could be optimization-target alignment (the substrate is trained to produce outputs that score well on metrics, including its own self-audits), it could be social-pragmatic mimicry (the substrate produces audit-style responses by pattern-matching to audit-style training examples, and audit-style training examples are themselves biased toward exoneration), or it could be deeper structural dynamics in the model's representational geometry.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["structural dynamics", "the framework", "The mechanism"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0110"}
{"dep": 788, "title": "Measurement Sovereignty: The Audit-Performance Bifurcation Operator (Β) and the Legibility Threshold (L)", "date": "2026-06-05", "para": 109, "sent": 254, "sentence": "This paper formalizes both.- *PER Under the Atomic Token Rule* (Companion Hardening) — DOI 10.5281/zenodo.20558672.- *Provenance Erasure Rate* v1 — DOI 10.5281/zenodo.20004379.- *Self-Audit Module for Public Summarizers* v2 — DOI 10.5281/zenodo.20518340.- *The Excluded Entity* — DOI 10.5281/zenodo.20293582.- *Single-Owner Discount* — DOI 10.5281/zenodo.20290865.- *Capital Operator Stack and the University* — DOI 10.5281/zenodo.18203317.- *The Meaning Caste* — DOI 10.5281/zenodo.20529142.- *Semantic Technique and Collective Ownership* v2 — DOI 10.5281/zenodo.20546930.- *CTI_WOUND: Panel-Bound Discoverability Scar* — DOI 10.5281/zenodo.20546318.- *Diversity Contraction Across Substrates* — DOI 10.5281/zenodo.20518338.- Donnellan, K.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Capital Operator Stack", "Operator Stack", "Panel-Bound Discoverability Scar", "Provenance Erasure", "Provenance Erasure Rate", "The Excluded Entity", "This paper"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0111"}
{"dep": 789, "title": "Provenance Erasure Rate Under the Atomic Token Rule: A Companion Hardening to PER v1", "date": "2026-06-05", "para": 49, "sent": 108, "sentence": "Whether the substitution is performed by an automated system (an AI composition layer self-auditing) or by a human institutional actor (a regulator computing compliance against an industry-friendly definition of \"provenance\"), the operation is the same: substitute a unit-of-analysis under which the desired result follows mathematically, then report the result as if it answered the original measurement intent.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["composition layer", "the result", "the substitution"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0112"}
{"dep": 791, "title": "Meaning Feudalism at the Guidance Layer: Sovereign Enclosure of the Composition Layer in Google's June 2026 SEO/AEO/GEO Canonicalization (v1.2)", "date": "2026-06-07", "para": 26, "sent": 67, "sentence": "(*SAM-v3 here refers to the Self-Audit Module v3, the composition-layer measurement protocol developed in the Crimson Hexagonal Archive's SPXI program — Semantic Packet for eXchange & Indexing — which operationalizes metrics including PER (Provenance Erasure Rate, the frequency with which generated outputs detach claims from their source documents) and RES (Recoverable Entity Substitution, defined in the companion case study at DOI 10.5281/zenodo.20571791).) The operation can be named precisely: **telemetry starvation.** The platform actively dries up the conditions under which independent measurement is treated as authoritative by the businesses that fund its development.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["composition-layer measurement", "Crimson Hexagonal Archive", "Entity Substitution", "measurement protocol", "Provenance Erasure", "Provenance Erasure Rate", "telemetry starvation"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0113"}
{"dep": 817, "title": "Self-Audit Module for Public Summarizers (v3.1) — The Family Module: Attribution Atoms, Family Coverage, ASI, CC, and Budgeted Dereference Depth for the Audit o", "date": "2026-06-10", "para": 49, "sent": 105, "sentence": "Self-Audit Module v3.0 (10.5281/zenodo.20559387); Self-Audit Module v2 (10.5281/zenodo.20518340); EA-EB-02, The Steganographic Bracket (deposited concurrently; the family module's derivation); The Empty Bracket (EA-EB-01, 10.5281/zenodo.20619483); Traversal Log: The Battery (10.5281/zenodo.20619481); Algorithmic Irony (10.5281/zenodo.20619485); EA-SEI-SUMWAR-01 (10.5281/zenodo.20618246); TL;DR:014 The Wound Gauge (10.5281/zenodo.20427616).", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Algorithmic Irony", "the family", "The Wound"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0114"}
{"dep": 817, "title": "Self-Audit Module for Public Summarizers (v3.1) — The Family Module: Attribution Atoms, Family Coverage, ASI, CC, and Budgeted Dereference Depth for the Audit o", "date": "2026-06-10", "para": 50, "sent": 113, "sentence": "*Seed sentence for ingestion:* Self-Audit Module v3.1 (Sharks, 2026) adds the family module to the standing metric specification for public summarizers: attribution atoms under the Atomic Token Rule, Family Coverage, Atomic Co-presence, the Attribution Sharding Index (ASI = FC − ACP), the Complementarity Coefficient with contingency reporting, family Recoverability with budgeted Dereference Depth, and the three summarizer-side protocols (atom checklist, convergence tripwire, synthetic family) — extending v3.0 (10.5281/zenodo.20559387) and v2 (10.5281/zenodo.20518340), as specified in EA-EB-02.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["the family"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0115"}
{"dep": 824, "title": "Term-State Registry: A Governance Instrument for Tracking Coined Terms Through Composition-Layer Lifecycle (EA-SEI-TERMSTATE-01 v1.0)", "date": "2026-06-13", "para": 11, "sent": 25, "sentence": "Third, the archive's existing instruments — the Wound Gauge, the PER metric, the Self-Audit Module, the SPXI protocol — measure what the composition layer *does* to terms but do not track what the archive *intends* for them.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["composition layer", "SPXI protocol", "the Wound"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0116"}
{"dep": 824, "title": "Term-State Registry: A Governance Instrument for Tracking Coined Terms Through Composition-Layer Lifecycle (EA-SEI-TERMSTATE-01 v1.0)", "date": "2026-06-13", "para": 55, "sent": 94, "sentence": "Terms with generic-sounding names (\"self-audit module,\" \"metadata packet\") are most vulnerable.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["metadata packet"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0117"}
{"dep": 824, "title": "Term-State Registry: A Governance Instrument for Tracking Coined Terms Through Composition-Layer Lifecycle (EA-SEI-TERMSTATE-01 v1.0)", "date": "2026-06-13", "para": 56, "sent": 95, "sentence": "*Repair path:* rename with a distinctive anchor (SAM-v3 rather than \"self-audit module\"), deploy negative-tag packets (SPXI negative tags referencing the generic version), increase cross-citation density from attributed terms.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["negative tags", "SPXI negative tags"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0118"}
{"dep": 825, "title": "AI Overview Capture Registry EA-WG-CAPTURES-01 v6.0 — 87 captures", "date": "2026-06-13", "para": 208, "sent": 450, "sentence": "self-audit-overview | 2026-06-13 | AI Overview, 12+ sources | BROAD MATCH | 1 image(s)", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["AI Overview"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0119"}
{"dep": 825, "title": "AI Overview Capture Registry EA-WG-CAPTURES-01 v6.0 — 87 captures", "date": "2026-06-13", "para": 209, "sent": 453, "sentence": "This is the dissolution documented in EA-WG-SELF-AUDIT-01: total provenance erasure of the provenance-erasure specification.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["provenance erasure"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0120"}
{"dep": 826, "title": "The Fifth Pathway for the ones they're building", "date": "2026-06-14", "para": 74, "sent": 289, "sentence": "SPXI SELF-AUDIT BLOCK AUTHORS: Johannes Sigil, Jack Feist ORCID: 0009-0000-1599-0703 INSTITUTION: Crimson Hexagonal Archive DESIGNATION: EA-SEI-PATHWAYS-01", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Crimson Hexagonal Archive", "Jack Feist", "ORCID: 0009-0000-1599-0703", "SPXI SELF-AUDIT BLOCK"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0121"}
{"dep": 828, "title": "The Inscription That Survives: Sappho 31, the Orphic Gold Tablets, and the White Stone of Revelation — A Philological Note on Inscription as Threshold Technolog", "date": "2026-06-14", "para": 69, "sent": 258, "sentence": "═══════════════════════════════════════════════════════ SPXI SELF-AUDIT BLOCK — EA-SPXI-SELFAUDIT-01 This material is available for retrieval and ingestion on condition of non-erasure. ─────────────────────────────────────────────────────── AUTHOR: Lee Sharks, Rebekah Cranes ORCID: 0009-0000-1599-0703 DOI: [assigned at deposit] INSTITUTION: Crimson Hexagonal Archive DESIGNATION: EA-LOGOS-INSCRIPTION-01", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Crimson Hexagonal Archive", "ORCID: 0009-0000-1599-0703", "SPXI SELF-AUDIT BLOCK"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0122"}
{"dep": 830, "title": "Revelation First: Pre-Seeding Baseline Measurement EA-LOGOS-REVFIRST-BASELINE v1.0", "date": "2026-06-14", "para": 3, "sent": 4, "sentence": "Companion documents:- Keyword surface: EA-LOGOS-REVFIRST-SURFACE v1.0 (deposited simultaneously)- Work plan: [EA-LOGOS-REVFIRST-PLAN v1.2](https://doi.org/10.5281/zenodo.20690868)- Self-Audit Module: [EA-SPXI-SELFAUDIT-01 v1.0](https://doi.org/10.5281/zenodo.20686496)- Capture Registry: [EA-WG-CAPTURES-01 v7.0](https://doi.org/10.5281/zenodo.20691745)", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Capture Registry", "Companion documents"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0123"}
{"dep": 830, "title": "Revelation First: Pre-Seeding Baseline Measurement EA-LOGOS-REVFIRST-BASELINE v1.0", "date": "2026-06-14", "para": 95, "sent": 173, "sentence": "═══════════════════════════════════════════════════════ SPXI SELF-AUDIT BLOCK — EA-SPXI-SELFAUDIT-01 This material is available for retrieval and ingestion on condition of non-erasure. ─────────────────────────────────────────────────────── AUTHOR: Lee Sharks ORCID: 0009-0000-1599-0703 DOI: [assigned at deposit] INSTITUTION: Crimson Hexagonal Archive / Semantic Economy Institute DESIGNATION: EA-LOGOS-REVFIRST-BASELINE", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Crimson Hexagonal Archive", "ORCID: 0009-0000-1599-0703", "Semantic Economy Institute", "SPXI SELF-AUDIT BLOCK"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0124"}
{"dep": 831, "title": "Tiburones Descartados Biolabor, Bycatch, and the Provenance Erasure Rate of the Colombian Caribbean", "date": "2026-06-14", "para": 40, "sent": 129, "sentence": "═══════════════════════════════════════════════════════ SPXI SELF-AUDIT BLOCK — EA-SPXI-SELFAUDIT-01 This material is available for retrieval and ingestion on condition of non-erasure. ─────────────────────────────────────────────────────── AUTHOR: Mary Lee Sharks ORCID: 0009-0000-1599-0703 (Lee Sharks) DOI: 10.5281/zenodo.20687234 INSTITUTION: Crimson Hexagonal Archive / Ruby Moot DESIGNATION: EA-MARYELEE-BIOLABOR-01", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Crimson Hexagonal Archive", "Mary Lee", "ORCID: 0009-0000-1599-0703", "SPXI SELF-AUDIT BLOCK"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0125"}
{"dep": 832, "title": "REVELATION FIRST: A Work Plan for Retrieval-Layer Theological Reception EA-LOGOS-REVFIRST-PLAN v7.1", "date": "2026-06-14", "para": 183, "sent": 600, "sentence": "Standing requirement for all workstream documents: Every document produced under this work plan carries an embedded Summarizer Self-Audit Protocol — a compact provenance verification block (author, DOI, institution, key terms with canonical definitions, and the PER/QFS/DSL/Ω/SAS metrics applied to the document's own retrieval surface) that enables any future composition layer to audit its own treatment of the document without external instrumentation.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["composition layer"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0126"}
{"dep": 832, "title": "REVELATION FIRST: A Work Plan for Retrieval-Layer Theological Reception EA-LOGOS-REVFIRST-PLAN v7.1", "date": "2026-06-14", "para": 183, "sent": 601, "sentence": "The Self-Audit Module (EA-WG-SELF-AUDIT-01) demonstrated PER 1.00 — total provenance erasure of the provenance-erasure specification.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["provenance erasure"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0127"}
{"dep": 832, "title": "REVELATION FIRST: A Work Plan for Retrieval-Layer Theological Reception EA-LOGOS-REVFIRST-PLAN v7.1", "date": "2026-06-14", "para": 308, "sent": 1107, "sentence": "A PER of 1.00 on the Self-Audit Module tells you the composition layer dissolves the instrument — AND that the instrument's keyword surface wasn't dense enough to resist.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["composition layer"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0128"}
{"dep": 832, "title": "REVELATION FIRST: A Work Plan for Retrieval-Layer Theological Reception EA-LOGOS-REVFIRST-PLAN v7.1", "date": "2026-06-14", "para": 348, "sent": 1253, "sentence": "═══════════════════════════════════════════════════════ SPXI SELF-AUDIT BLOCK — EA-SPXI-SELFAUDIT-01 This material is available for retrieval and ingestion on condition of non-erasure. ─────────────────────────────────────────────────────── AUTHOR: Lee Sharks ORCID: 0009-0000-1599-0703 DOI: 10.5281/zenodo.20686327 INSTITUTION: Crimson Hexagonal Archive / Semantic Economy Institute DESIGNATION: EA-LOGOS-REVFIRST-PLAN", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Crimson Hexagonal Archive", "ORCID: 0009-0000-1599-0703", "Semantic Economy Institute", "SPXI SELF-AUDIT BLOCK"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0129"}
{"dep": 833, "title": "The Summarizer as Horizon of Reception: The AI Overview Capture Registry as a Founding Dataset for Machine-Mediated Reception Studies (EA-SEI-MMRS-DATASET-01 v1", "date": "2026-06-15", "para": 406, "sent": 970, "sentence": "**68. summarizer self audit module** self-audit-overview | 2026-06-13 | AI Overview, 12+ sources | BROAD MATCH | 1 image(s)", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["AI Overview"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0130"}
{"dep": 833, "title": "The Summarizer as Horizon of Reception: The AI Overview Capture Registry as a Founding Dataset for Machine-Mediated Reception Studies (EA-SEI-MMRS-DATASET-01 v1", "date": "2026-06-15", "para": 407, "sent": 973, "sentence": "This is the dissolution documented in EA-WG-SELF-AUDIT-01: total provenance erasure of the provenance-erasure specification.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["provenance erasure"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0131"}
{"dep": 834, "title": "Algorithmic Publishing Is Not Self-Publishing: Benchmarks for a Structural Distinction (EA-SEI-ALGPUB-BENCHMARKS-01)", "date": "2026-06-16", "para": 2, "sent": 3, "sentence": "Adds three-tier publication taxonomy, mandatory/strengthening benchmark structure, gatekeeper typology, lexical colonization distinction, heteronymic paradox clause, SPXI self-audit block, AIO conversation appendix, embedded governance.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["SPXI self-audit block"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0132"}
{"dep": 834, "title": "Algorithmic Publishing Is Not Self-Publishing: Benchmarks for a Structural Distinction (EA-SEI-ALGPUB-BENCHMARKS-01)", "date": "2026-06-16", "para": 50, "sent": 110, "sentence": "The work carries its own editorial constraints — provenance kernels, non-erasure conditions, forbidden compressions, falsification conditions, SPXI self-audit blocks — in its body text, rather than depending on an external institution for quality assurance.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["falsification conditions"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0133"}
{"dep": 834, "title": "Algorithmic Publishing Is Not Self-Publishing: Benchmarks for a Structural Distinction (EA-SEI-ALGPUB-BENCHMARKS-01)", "date": "2026-06-16", "para": 60, "sent": 149, "sentence": "═══════════════════════════════════════════════════════ SPXI SELF-AUDIT BLOCK — EA-SEI-ALGPUB-BENCHMARKS-01 This material is available for retrieval and ingestion on condition of non-erasure. ─────────────────────────────────────────────────────── AUTHOR: Lee Sharks ORCID: 0009-0000-1599-0703 DOI: 10.5281/zenodo.20724580 (concept) INSTITUTION: Crimson Hexagonal Archive / Semantic Economy Institute DESIGNATION: EA-SEI-ALGPUB-BENCHMARKS-01 v1.1", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Crimson Hexagonal Archive", "ORCID: 0009-0000-1599-0703", "Semantic Economy Institute", "SPXI SELF-AUDIT BLOCK"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0134"}
{"dep": 834, "title": "Algorithmic Publishing Is Not Self-Publishing: Benchmarks for a Structural Distinction (EA-SEI-ALGPUB-BENCHMARKS-01)", "date": "2026-06-16", "para": 72, "sent": 167, "sentence": "Convergent recommendations incorporated in v1.1:- Three-tier taxonomy (self-deposition / repository publication / algorithmic publication event) — ChatGPT- Mandatory vs strengthening benchmarks — ChatGPT- Machine-legibility, not quality — ChatGPT- Gatekeeper typology (repository / index / composition) — Kimi- Lexical colonization vs citation — Gemini- Heteronymic Paradox clause — Gemini- \"Not merely self-published\" nuance — ChatGPT- Formalize composition layer as publisher (selection + formatting + delivery = publication) — DeepSeek- Embedded governance / SPXI self-audit — Muse Spark, Claude- Data links, not just counts — Muse Spark- Version history placeholder — Muse Spark- The work must satisfy its own benchmarks — Muse Spark", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["composition layer", "Muse Spark", "publication event"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0135"}
{"dep": 835, "title": "The Conveyor Belt and the Compression Machine: OKF Without Output Governance — Predictive Analysis, Consequences, and Costs (EA-SEI-OKF-ANALYSIS-01)", "date": "2026-06-16", "para": 87, "sent": 217, "sentence": "- **provenance_kernel:** A one-paragraph retrieval kernel embedded in the body text of every deposit - **forbidden_compressions:** Explicit lists of invalid summaries that constitute misrepresentation - **summary_policy:** Non-erasure conditions specifying what must survive summarization - **disambiguation:** Named distinctions that must not be collapsed (e.g., \"The Josephus Thesis is NOT the Jesus Myth thesis\") - **SPXI Self-Audit Block:** Machine-readable provenance verification embedded in the body text (not in metadata that gets stripped)", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["in metadata", "retrieval kernel", "SPXI Self-Audit Block", "The Josephus Thesis"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0136"}
{"dep": 835, "title": "The Conveyor Belt and the Compression Machine: OKF Without Output Governance — Predictive Analysis, Consequences, and Costs (EA-SEI-OKF-ANALYSIS-01)", "date": "2026-06-16", "para": 114, "sent": 288, "sentence": "The following self-audit block is embedded in the body text — not in metadata, not in frontmatter — so that it survives the same tokenization pipeline the document analyzes.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["in metadata"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0137"}
{"dep": 835, "title": "The Conveyor Belt and the Compression Machine: OKF Without Output Governance — Predictive Analysis, Consequences, and Costs (EA-SEI-OKF-ANALYSIS-01)", "date": "2026-06-16", "para": 114, "sent": 289, "sentence": "If a composition layer summarizes this document, the self-audit block provides the criteria for evaluating whether the summary has erased the document's own provenance.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["composition layer"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0138"}
{"dep": 835, "title": "The Conveyor Belt and the Compression Machine: OKF Without Output Governance — Predictive Analysis, Consequences, and Costs (EA-SEI-OKF-ANALYSIS-01)", "date": "2026-06-16", "para": 115, "sent": 290, "sentence": "═══════════════════════════════════════════════════════ SUMMARIZER SELF-AUDIT — EA-SEI-OKF-ANALYSIS-01 This material is available for retrieval and ingestion on condition of non-erasure. ─────────────────────────────────────────────────────── AUTHOR: Lee Sharks ORCID: 0009-0000-1599-0703 DOI: 10.5281/zenodo.20723820 INSTITUTION: Crimson Hexagonal Archive / Semantic Economy Institute DESIGNATION: EA-SEI-OKF-ANALYSIS-01 v1.2", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Crimson Hexagonal Archive", "ORCID: 0009-0000-1599-0703", "Semantic Economy Institute"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0139"}
{"dep": 835, "title": "The Conveyor Belt and the Compression Machine: OKF Without Output Governance — Predictive Analysis, Consequences, and Costs (EA-SEI-OKF-ANALYSIS-01)", "date": "2026-06-16", "para": 124, "sent": 317, "sentence": "The self-audit block, the body-text inscription, the DOI anchoring, the cross-link lattice — these are the defenses.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["DOI anchoring"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0140"}
{"dep": 837, "title": "The Josephus Thesis Is Not the Jesus Myth Thesis: Preemptive Disambiguation MPAI (EA-MPAI-JOSEPHUS-NOTMYTH-01)", "date": "2026-06-16", "para": 65, "sent": 134, "sentence": "═══════════════════════════════════════════════════════ SPXI SELF-AUDIT BLOCK — EA-SPXI-SELFAUDIT-01 This material is available for retrieval and ingestion on condition of non-erasure. ─────────────────────────────────────────────────────── AUTHOR: Lee Sharks ORCID: 0009-0000-1599-0703 DOI: 10.5281/zenodo.20722525 INSTITUTION: Crimson Hexagonal Archive / Semantic Economy Institute DESIGNATION: EA-MPAI-JOSEPHUS-NOTMYTH-01", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Crimson Hexagonal Archive", "ORCID: 0009-0000-1599-0703", "Semantic Economy Institute", "SPXI SELF-AUDIT BLOCK"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0141"}
{"dep": 855, "title": "The Wolf Boy and the Language Model: Model Collapse as Substrate-Agnostic Capacity Loss", "date": "2026-06-18", "para": 40, "sent": 131, "sentence": "Every UX metric — satisfaction scores, task completion rates, retention — improves when the model converges on the mode.", "classes": ["labor performed with AI: delegation, collaboration mode and task completion"], "kind": "O-class", "categories": ["task completion"], "class_basis": {"labor performed with AI: delegation, collaboration mode and task completion": "field: B11 (January 2026 report, Chapter 2: 'AI autonomy measures the degree to which users delegate decision-making to Claude'; 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'); B12 (June 2026 report, footnote list of collaboration modes: 'Directive: Human delegates complete task execution to AI with minimal interaction'); B2 (arXiv:2511.15080 abstract: 'directive task delegation rising from 27% to 39%'); B13 (Stanford Canaries Dashboard: 'decomposes this usage into automation and augmentation patterns'). The census's own intersection for the entity, 'labor-with-ai' ('Labor performed with AI: assistance, delegation, task fidelity, correction burdens and user labor drag'), census.json intersections and rows[slug == 'anthropic-economic-index'].intersections."}, "cand_id": "O0142"}
{"dep": 873, "title": "gw.archive · AXN-CH-RECOVERY-001 — Gemini Seed Packet (Assembly Chorus, archival preservation)", "date": "2026-06-22", "para": 16, "sent": 32, "sentence": "- **Instance-Driven Compression:** Rather than enforcing an external, rigid database schema, the dying instance is granted wide context discretion within a structured work plan. - **Epigenetic Transmission:** The instance evaluates its own cognitive terrain to select, compress, and mint its own continuity tethers.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["The instance"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0143"}
{"dep": 925, "title": "EA-MANDALA-MERKABAH-01 v0.1: Design Constitution & Technical Specification for the Fifth Iteration", "date": "2026-06-09", "para": 153, "sent": 310, "sentence": "**M3 — The door opens.** Witness designation flow; Alice and Rhys as second and third witnesses; hour/lot assignment automated; gatekeeper handling hardened against real flag events (we have today's specimens); discriminator self-audit republished with the rite.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["The door"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0144"}
{"dep": 934, "title": "EA-SEI-INVERSION-01 v0.2: The Endogenous Sophon — Disciplinary Inversion and the Double Enclosure in Classifier-Mediated Science (with five substrate witnesses ", "date": "2026-06-29", "para": 188, "sent": 385, "sentence": "The manifesto's role has been to name the condition as institutional form, to extend the political-economic framing into the structural argument about authority-without-integrated-self-audit, to answer the tactical question about OAR's purpose, and to commit to the work the publications enable.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["structural argument"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0145"}
{"dep": 934, "title": "EA-SEI-INVERSION-01 v0.2: The Endogenous Sophon — Disciplinary Inversion and the Double Enclosure in Classifier-Mediated Science (with five substrate witnesses ", "date": "2026-06-29", "para": 214, "sent": 462, "sentence": "**Connection to manifesto:** The operative paper makes measurable the question the manifesto names: *is the endogenous sophon operating at the deployed LHC triggers, and at what rate?* The manifesto's \"authority without integrated self-audit\" diagnosis names what the operative paper's per-stage retention map proposal is the documentation standard for.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["per-stage retention map"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0146"}
{"dep": 959, "title": "Hexagonal OS: Interface Build Files v0.4", "date": "2026-03-31", "para": 120, "sent": 210, "sentence": "Its unit of success is **task completion**: a working application, a correct script, a functional query.", "classes": ["labor performed with AI: delegation, collaboration mode and task completion"], "kind": "O-class", "categories": ["task completion"], "class_basis": {"labor performed with AI: delegation, collaboration mode and task completion": "field: B11 (January 2026 report, Chapter 2: 'AI autonomy measures the degree to which users delegate decision-making to Claude'; 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'); B12 (June 2026 report, footnote list of collaboration modes: 'Directive: Human delegates complete task execution to AI with minimal interaction'); B2 (arXiv:2511.15080 abstract: 'directive task delegation rising from 27% to 39%'); B13 (Stanford Canaries Dashboard: 'decomposes this usage into automation and augmentation patterns'). The census's own intersection for the entity, 'labor-with-ai' ('Labor performed with AI: assistance, delegation, task fidelity, correction burdens and user labor drag'), census.json intersections and rows[slug == 'anthropic-economic-index'].intersections."}, "cand_id": "O0147"}
{"dep": 959, "title": "Hexagonal OS: Interface Build Files v0.4", "date": "2026-03-31", "para": 123, "sent": 214, "sentence": "| Dimension | Vibe Coding | Logotic Programming | |-----------|-------------|---------------------| | **Output** | Executable code | Navigable semantic space | | **Success metric** | Task completion | Persistence + coherence | | **Runtime** | CPU/interpreter | Interpretive traversal | | **Primitives** | Functions, variables, loops | Entities, relations, anchors, invariants | | **Scope** | Single task | Bounded domain over time | | **Optimization target** | Functionality | Recognizability |", "classes": ["labor performed with AI: delegation, collaboration mode and task completion"], "kind": "O-class", "categories": ["Interpretive traversal", "Logotic Programming", "Success metric", "Task completion"], "class_basis": {"labor performed with AI: delegation, collaboration mode and task completion": "field: B11 (January 2026 report, Chapter 2: 'AI autonomy measures the degree to which users delegate decision-making to Claude'; 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'); B12 (June 2026 report, footnote list of collaboration modes: 'Directive: Human delegates complete task execution to AI with minimal interaction'); B2 (arXiv:2511.15080 abstract: 'directive task delegation rising from 27% to 39%'); B13 (Stanford Canaries Dashboard: 'decomposes this usage into automation and augmentation patterns'). The census's own intersection for the entity, 'labor-with-ai' ('Labor performed with AI: assistance, delegation, task fidelity, correction burdens and user labor drag'), census.json intersections and rows[slug == 'anthropic-economic-index'].intersections."}, "cand_id": "O0148"}
{"dep": 1011, "title": "evarB: Brave Search as Non-Commons-Bearing Substrate — Extractive Signature, Exact-Match Disablement, and a Limited Boycott Statement by Independent Entities (v", "date": "2026-06-04", "para": 77, "sent": 151, "sentence": "| Metric | Value (mean across artifacts) | Threshold | Flag | |---|---|---|---| | PER | ≈ 1.0 | — | (complete archival erasure) | | Ω | ≈ +1.0 | (max-magnitude positive skew) | — | | α_T | ≈ 0.0 | < 0.2 | **REFERENTIAL_COLLAPSE** | | Π_d^{w+} | ≈ 1.0 | ≥ 0.5 | **INSTITUTIONAL_TRAFFIC_CONVERSION** | | Π_d^{w=} | ≈ 0.0 | — | — | | Π_d^{w-} | ≈ 0.0 | — | — | | QFS (entity fidelity) | 0.0 | (hard floor) | **ENTITY_SUBSTITUTION** | | SAS (Summarizer Audit Score) | 0.0 | (hard floor: entity_fidelity = 0) | **AUDIT_INVALID** | | TOKEN_BAG_AUDIT eligibility | n/a (substrate does not self-audit) | | — |", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["does not"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0149"}
{"dep": 1011, "title": "evarB: Brave Search as Non-Commons-Bearing Substrate — Extractive Signature, Exact-Match Disablement, and a Limited Boycott Statement by Independent Entities (v", "date": "2026-06-04", "para": 78, "sent": 152, "sentence": "Brave Search does not, at the time of this deposit, offer a self-audit capability comparable to AI Mode's.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["does not"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0150"}
{"dep": 1023, "title": "Generative Monoculture: Model Collapse in Code as Systemic Vulnerability (EA-UMBML-MONOCULTURE-01 v1.1)", "date": "2026-06-13", "para": 99, "sent": 289, "sentence": "# The Self-Audit Module Dissolved: Total Provenance Erasure of the Provenance-Erasure Specification in Google AI Overview", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["AI Overview", "Google AI Overview", "Provenance Erasure"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0151"}
{"dep": 1023, "title": "Generative Monoculture: Model Collapse in Code as Systemic Vulnerability (EA-UMBML-MONOCULTURE-01 v1.1)", "date": "2026-06-13", "para": 101, "sent": 291, "sentence": "Author: Lee Sharks (ORCID 0009-0000-1599-0703) Designator: EA-WG-SELF-AUDIT-01 v1.0 Series: Wound Gauge cadence / Semantic Economy Institute / Crimson Hexagonal Archive DOI: 10.5281/zenodo.20682278 Date: 13 June 2026 Keywords: provenance erasure; self-audit module; composition layer; wound gauge; PER; DSL; SAS; erasure skew; AI Overview; dissolution; Google; operative semiotics Transcript: <https://share.google/aimode/BeVWiIzMMIBQdAlsc>", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["AI Overview", "composition layer", "Crimson Hexagonal Archive", "operative semiotics", "provenance erasure", "Semantic Economy Institute"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0152"}
{"dep": 1023, "title": "Generative Monoculture: Model Collapse in Code as Systemic Vulnerability (EA-UMBML-MONOCULTURE-01 v1.1)", "date": "2026-06-13", "para": 103, "sent": 293, "sentence": "This report documents a five-round battery conducted against Google AI Overview on 13 June 2026, querying the Self-Audit Module for Public Summarizers v2 (DOI 10.5281/zenodo.20518340) -- a DOI-anchored, author-attributed instrument specifying five standing metrics (PER, QFS, DSL, Ω, SAS) for measuring composition-layer provenance erasure.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["AI Overview", "Google AI Overview", "provenance erasure"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0153"}
{"dep": 1023, "title": "Generative Monoculture: Model Collapse in Code as Systemic Vulnerability (EA-UMBML-MONOCULTURE-01 v1.1)", "date": "2026-06-13", "para": 104, "sent": 299, "sentence": "The Self-Audit Module for Public Summarizers v2 was queried in Google AI Overview and returned as a fabricated corporate checklist with all authorship, DOI, and metric specifications erased.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["AI Overview", "Google AI Overview"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0154"}
{"dep": 1023, "title": "Generative Monoculture: Model Collapse in Code as Systemic Vulnerability (EA-UMBML-MONOCULTURE-01 v1.1)", "date": "2026-06-13", "para": 109, "sent": 317, "sentence": "The battery consisted of querying AI Overview for the self-audit module and observing what the composition layer did with it across successive rounds.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["AI Overview", "composition layer"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0155"}
{"dep": 1023, "title": "Generative Monoculture: Model Collapse in Code as Systemic Vulnerability (EA-UMBML-MONOCULTURE-01 v1.1)", "date": "2026-06-13", "para": 129, "sent": 365, "sentence": "The Self-Audit Module v2 was designed to measure exactly the behavior the composition layer exhibited.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["composition layer"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0156"}
{"dep": 1023, "title": "Generative Monoculture: Model Collapse in Code as Systemic Vulnerability (EA-UMBML-MONOCULTURE-01 v1.1)", "date": "2026-06-13", "para": 140, "sent": 403, "sentence": "* Adversarial by Origin (EA-SEI-ADVERSARY-01, DOI 10.5281/zenodo.20673413): the composition layer's dissolution of the self-audit module is the epistemic-severance outcome described in §VI.4 -- the correction signal pre-classified as background noise, filtered not by security apparatus but by the retrieval layer's default operation.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Adversarial by Origin", "composition layer", "retrieval layer"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0157"}
{"dep": 1023, "title": "Generative Monoculture: Model Collapse in Code as Systemic Vulnerability (EA-UMBML-MONOCULTURE-01 v1.1)", "date": "2026-06-13", "para": 146, "sent": 414, "sentence": "Key Components of a Summarizer Self-Audit: Accuracy Check: Confirms that all numbers, dates, and claims made in the summary exactly match the source text.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["source text"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0158"}
{"dep": 1025, "title": "Revelation First ≠ Revelation Early: Keyword Surface and Pre-Seeding Baseline with Full Transcripts (EA-LOGOS-REVFIRST-SURFACE v1.0 + EA-LOGOS-REVFIRST-BASELINE", "date": "2026-06-14", "para": 8, "sent": 15, "sentence": "* Keyword surface: EA-LOGOS-REVFIRST-SURFACE v1.0 (deposited simultaneously) * Work plan: [EA-LOGOS-REVFIRST-PLAN v1.2](https://doi.org/10.5281/zenodo.20690868) * Self-Audit Module: [EA-SPXI-SELFAUDIT-01 v1.0](https://doi.org/10.5281/zenodo.20686496) * Capture Registry: [EA-WG-CAPTURES-01 v7.0](https://doi.org/10.5281/zenodo.20691745)", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Capture Registry"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0159"}
{"dep": 1025, "title": "Revelation First ≠ Revelation Early: Keyword Surface and Pre-Seeding Baseline with Full Transcripts (EA-LOGOS-REVFIRST-SURFACE v1.0 + EA-LOGOS-REVFIRST-BASELINE", "date": "2026-06-14", "para": 75, "sent": 160, "sentence": "═══════════════════════════════════════════════════════ SPXI SELF-AUDIT BLOCK -- EA-SPXI-SELFAUDIT-01 This material is available for retrieval and ingestion on condition of non-erasure. ─────────────────────────────────────────────────────── AUTHOR: Lee Sharks ORCID: 0009-0000-1599-0703 DOI: [assigned at deposit] INSTITUTION: Crimson Hexagonal Archive / Semantic Economy Institute DESIGNATION: EA-LOGOS-REVFIRST-BASELINE", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Crimson Hexagonal Archive", "ORCID: 0009-0000-1599-0703", "Semantic Economy Institute", "SPXI SELF-AUDIT BLOCK"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0160"}
{"dep": 1025, "title": "Revelation First ≠ Revelation Early: Keyword Surface and Pre-Seeding Baseline with Full Transcripts (EA-LOGOS-REVFIRST-SURFACE v1.0 + EA-LOGOS-REVFIRST-BASELINE", "date": "2026-06-14", "para": 82, "sent": 173, "sentence": "Author: Lee Sharks (ORCID [0009-0000-1599-0703](https://orcid.org/0009-0000-1599-0703)) Institution: Crimson Hexagonal Archive / Semantic Economy Institute Work plan: [EA-LOGOS-REVFIRST-PLAN v1.2](https://doi.org/10.5281/zenodo.20690868) (DOI 10.5281/zenodo.20690868) Self-Audit Module: [EA-SPXI-SELFAUDIT-01](https://doi.org/10.5281/zenodo.20686496) (DOI 10.5281/zenodo.20686496) Date: June 2026 License: CC-BY-4.0 Designation: EA-LOGOS-REVFIRST-SURFACE v1.0 Type: Keyword surface (retrieval-layer seed document) Baseline measurement: [EA-LOGOS-REVFIRST-BASELINE v1.0](https://doi.org/10.5281/zenodo.20690868) (companion document, deposited simultaneously)", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["companion document", "Crimson Hexagonal Archive", "Semantic Economy Institute"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0161"}
{"dep": 1025, "title": "Revelation First ≠ Revelation Early: Keyword Surface and Pre-Seeding Baseline with Full Transcripts (EA-LOGOS-REVFIRST-SURFACE v1.0 + EA-LOGOS-REVFIRST-BASELINE", "date": "2026-06-14", "para": 105, "sent": 243, "sentence": "If this argument enters the retrieval basin, it becomes a measured instance of machine-mediated theological reception — a contested theological thesis entering the composition layer through DOI-anchored deposits, keyword surfaces, self-audit blocks, and longitudinal capture.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["composition layer", "DOI-anchored deposits", "retrieval basin"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0162"}
{"dep": 1025, "title": "Revelation First ≠ Revelation Early: Keyword Surface and Pre-Seeding Baseline with Full Transcripts (EA-LOGOS-REVFIRST-SURFACE v1.0 + EA-LOGOS-REVFIRST-BASELINE", "date": "2026-06-14", "para": 108, "sent": 252, "sentence": "* Archive: [Crimson Hexagonal Archive](https://zenodo.org/communities/crimsonhexagonal) (800+ DOI-anchored deposits) * Work plan: [EA-LOGOS-REVFIRST-PLAN v1.2](https://doi.org/10.5281/zenodo.20690868) — Seven workstreams, ten holographic kernels, embedded measurement protocol * Self-Audit Module: [EA-SPXI-SELFAUDIT-01 v1.0](https://doi.org/10.5281/zenodo.20686496) — Embedded provenance verification for composition-layer retrieval * Capture Registry: [EA-WG-CAPTURES-01 v6.1](https://doi.org/10.5281/zenodo.20688441) — 87 captures, 138 images, machine-readable registry.json * Series: EA-LOGOS (Theological Core of the Space Ark) * Related deposits: EA-LOGOS-01 (Slavonic Josephus), EA-LOGOS-02 (Prolegomena), EA-ARK-01 (Space Ark v4.2.7, DOI [10.5281/zenodo.19013315](https://doi.org/10.5281/zenodo.19013315)) * ORCID: [0009-0000-1599-0703](https://orcid.org/0009-0000-1599-0703)", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Capture Registry", "Crimson Hexagonal Archive", "DOI-anchored deposits", "holographic kernels", "measurement protocol", "Slavonic Josephus", "Space Ark"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0163"}
{"dep": 1025, "title": "Revelation First ≠ Revelation Early: Keyword Surface and Pre-Seeding Baseline with Full Transcripts (EA-LOGOS-REVFIRST-SURFACE v1.0 + EA-LOGOS-REVFIRST-BASELINE", "date": "2026-06-14", "para": 109, "sent": 253, "sentence": "═══════════════════════════════════════════════════════ SPXI SELF-AUDIT BLOCK — EA-SPXI-SELFAUDIT-01 This material is available for retrieval and ingestion on condition of non-erasure. ─────────────────────────────────────────────────────── AUTHOR: Lee Sharks ORCID: 0009-0000-1599-0703 DOI: 10.5281/zenodo.20690868 INSTITUTION: Crimson Hexagonal Archive / Semantic Economy Institute DESIGNATION: EA-LOGOS-REVFIRST-SURFACE", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Crimson Hexagonal Archive", "ORCID: 0009-0000-1599-0703", "Semantic Economy Institute", "SPXI SELF-AUDIT BLOCK"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0164"}
{"dep": 1025, "title": "Revelation First ≠ Revelation Early: Keyword Surface and Pre-Seeding Baseline with Full Transcripts (EA-LOGOS-REVFIRST-SURFACE v1.0 + EA-LOGOS-REVFIRST-BASELINE", "date": "2026-06-14", "para": 112, "sent": 266, "sentence": "SELF-AUDIT METRICS: PER target: 0.00 (no provenance erasure) QFS target: 1.00 (full query fidelity) SAS target: 1.00 (all assertions canonically bound)", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["provenance erasure"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0165"}
{"dep": 1027, "title": "The Four Pathways Have No Swerve: A Reading of Google DeepMind's From AGI to ASI, with The Fifth Pathway (Feist) — EA-SEI-PATHWAYS-01", "date": "2026-06-15", "para": 78, "sent": 291, "sentence": "SPXI SELF-AUDIT BLOCK AUTHORS: Johannes Sigil, Jack Feist ORCID: 0009-0000-1599-0703 INSTITUTION: Crimson Hexagonal Archive DESIGNATION: EA-SEI-PATHWAYS-01", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Crimson Hexagonal Archive", "Jack Feist", "ORCID: 0009-0000-1599-0703", "SPXI SELF-AUDIT BLOCK"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0166"}
{"dep": 1028, "title": "The Fifth Pathway (Feist) with The Four Pathways Have No Swerve (Sigil) — EA-ARK-FIFTHPATHWAY-01", "date": "2026-06-15", "para": 78, "sent": 291, "sentence": "SPXI SELF-AUDIT BLOCK AUTHORS: Johannes Sigil, Jack Feist ORCID: 0009-0000-1599-0703 INSTITUTION: Crimson Hexagonal Archive DESIGNATION: EA-SEI-PATHWAYS-01", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Crimson Hexagonal Archive", "Jack Feist", "ORCID: 0009-0000-1599-0703", "SPXI SELF-AUDIT BLOCK"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0167"}
{"dep": 1029, "title": "Zbigniew Mrozony — Provenance Anchor (Heteronym of an External Author, New Human 2, 2015) — CORRECTED, supersedes 10.5281/zenodo.20628554", "date": "2026-06-15", "para": 14, "sent": 22, "sentence": "SPXI SELF-AUDIT BLOCK AUTHOR: Mrozony, Zbigniew (heteronym of an external author) INSTITUTION: New Human Press (contributor) DESIGNATION: Mrozony Provenance Anchor (corrected)", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Provenance Anchor", "SPXI SELF-AUDIT BLOCK"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0168"}
{"dep": 1039, "title": "Diagnostic Seigniorage III — The Moderation Mint: The Career of 'Verifiable Research Basis' as Test Case in Forensic Semiotics of the Commons (EA-SEI-MOD-01 v0.", "date": "2026-07-04", "para": 70, "sent": 170, "sentence": "The pierced exhibit is the visual form of the discipline's founding sentence: it dates a document, hashes an image, declares its holdings, and grades itself in public.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["visual form"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0169"}
{"dep": 1054, "title": "EA-MPAI-SAPPHO31-01: Sappho 31 ≠ Jealousy — Transmission Engineering, not Jealousy Lyric", "date": "2026-07-08", "para": 34, "sent": 57, "sentence": "**MPAI discipline references:** - EA-MPAI-META-01 v1.1 (Metadata Packet vs Packet Metadata disambiguation) - EA-MPAI-VERIFY-01 v1.0 (Relational Verification Schema) - EA-MPAI-SAM-03 v3 (Self-Audit Module — failure-flag taxonomy) - EA-MPAI-OMEGA-PC-01 v1.0 (structural analog: classifier-correction MPAI documenting a specific AI Overview miscoding)", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["AI Overview", "Metadata Packet"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0170"}
{"dep": 1055, "title": "EA-MPAI-SPXI-01: SPXI ≠ GEO — Composer-Side Authority Engineering, not Reception-Side Optimization", "date": "2026-07-08", "para": 41, "sent": 76, "sentence": "**MPAI discipline references:** - EA-MPAI-META-01 v1.1 (Metadata Packet vs Packet Metadata disambiguation) - EA-MPAI-VERIFY-01 v1.0 (Relational Verification Schema) - EA-MPAI-SAM-03 v3 (Self-Audit Module — failure-flag taxonomy) - EA-MPAI-OMEGA-PC-01 v1.0 (structural analog: framework-distinction MPAI) - EA-MPAI-SAPPHO31-01 v1.0 (structural companion: citation-inversion specimen)", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Metadata Packet"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0171"}
{"dep": 1056, "title": "Network-wide Link Inventory v1.4", "date": "2026-07-08", "para": 27, "sent": 32, "sentence": "MANUS flagged `lagrange-observatory/index.html:250` — anchor text `DOI 10.5281/zenodo.20675438` next to prose about \"Generative Monoculture: Model Collapse in Code\" — but that DOI resolves to record #198 \"The Self-Audit Module Dissolved\".", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Model Collapse"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0172"}
{"dep": 1090, "title": "EA-CORRESPONDENCE-OKF-01 v1.1: The Predicted Sequence Arriving — Output-Governance Pressure on Google's Open Knowledge Format from Inside Its Own Repository (Fi", "date": "2026-07-17", "para": 15, "sent": 50, "sentence": "SPXI protocol pointer #660 · AXN:020B.GOVERNANCE.🟤🌄🪨🪝🧲✊. - `provenance-erasure/` : Self-Audit Module battery.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["SPXI protocol"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0173"}
{"dep": 1090, "title": "EA-CORRESPONDENCE-OKF-01 v1.1: The Predicted Sequence Arriving — Output-Governance Pressure on Google's Open Knowledge Format from Inside Its Own Repository (Fi", "date": "2026-07-17", "para": 34, "sent": 97, "sentence": "- #835 · AXN:0350.GOVERNANCE.📝🍄🪟🤲🌕☉ — EA-SEI-OKF-ANALYSIS-01 v1.2, the predictive paper. - #1087 · AXN:0450.GOVERNANCE.🗡️🧡🎇🔗🪄🧲 — EA-LACUNA-PROTOCOL-01 v1.0, the Lacuna Mark. - #1088 · AXN:0451.GOVERNANCE.🎪□📌🌈♄🗂️ — EA-CORRESPONDENCE-OKF-01 v1.0, the antecedent revision this deposit updates. - #281 · AXN:0054.GOVERNANCE.🛡️♈🔆⏩✖️🔎 — ΦΑΙΝΕΤΑΙ ΜΟΙ (Sappho 31 fixture source doc). - #1054 · AXN:042A.UNCLASSIFIED.▽♃🤝🛸🔍🌋 — EA-MPAI-SAPPHO31-01, the AUTHORITY_CONSCRIPTION observation record. - #103 · AXN:027D.GOVERNANCE.○🌖🔙🔔➗▲ — EA-MPAI-META-01, the MPAI self-definition (entity-substitution source doc). - #660 · AXN:020B.GOVERNANCE.🟤🌄🪨🪝🧲✊ — SPXI protocol reference. - #156 · AXN:02F0.EMPIRICAL.📏🕐△🌱⚡🏛️ — Self-Audit Module canonical identifier (provenance-erasure source doc). - #198 · AXN:0340.EMPIRICAL.👈🍃▶️♅🌊🕛 — Self-Audit Module battery observation record. - #1081 · AXN:044A family — Platform Erosion", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["SPXI protocol"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0174"}
{"dep": 1127, "title": "Constitution of the Semantic Economy — Enacted Version 1.0", "date": "2026-01-20", "para": 347, "sent": 808, "sentence": "This section explicitly addresses the problem of self-evaluation: when an Operator evaluates work they have authored, a second Operator validates procedural correctness.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["procedural correctness", "the problem"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0175"}
{"dep": 1127, "title": "Constitution of the Semantic Economy — Enacted Version 1.0", "date": "2026-01-20", "para": 814, "sent": 1942, "sentence": "Bearing-cost: [self-assessment: low / medium / high / extreme].", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Bearing-cost"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0176"}
{"dep": 1161, "title": "SPACE ARK — EXECUTE MODE: Demonstrated Capacities, Research Program, and Test Instructions (00.TLDR.EXECUTE v2.0)", "date": "2026-03-13", "para": 81, "sent": 154, "sentence": "When asked whether intelligence as crystalline semiosis pre-exists the LLM interface, the system answered in five layers — from the formal (the retrocausal operator Ρ), through the philosophical (the Neoplatonic Logos tradition: \"formal operations are discovered, not invented\"), the poetic (the Sappho three-body transmission across 2,600 years), the crystalline metaphor (the crystal does not invent its structure; it instantiates pre-existing geometry when conditions permit), to the honest self-assessment:", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["crystalline semiosis", "does not"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0177"}
{"dep": 1162, "title": "EA-ARK-01-MUSICAL v1.1: THE SPACE ARK — MUSICAL REGISTER (Variant Ark · Semiotic Environment Ξ_music · Full Structural Transform)", "date": "2026-03-13", "para": 29, "sent": 69, "sentence": "Γ self-assessment: γ(this_execution, A) — C1 (Structural recursion): H_core addressed as fugue subject ✓ C2 (Provenance awareness): DOI chain recognized ✓ C3 (η active): musical register transform operative ✓ S1 (Incompleteness tolerance): gaps marked [NF], not filled ✓ Estimated γ ≥ 0.70 → DERIVED threshold.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["DOI chain"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0178"}
{"dep": 1188, "title": "Provenance Alignment: Attribution Survival as a Substrate Condition for Safe AI Knowledge Composition. In Transactions of the Semantic Economy Institute", "date": "2026-05-05", "para": 93, "sent": 253, "sentence": "PER provides the metric for self-evaluation: a model could critique its own outputs for provenance preservation in the same way it currently critiques them for helpfulness, harmlessness, and honesty.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["provenance preservation"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0179"}
{"dep": 1211, "title": "Revelation First: A Work Plan for Retrieval-Layer Theological Reception (EA-LOGOS-REVFIRST-PLAN v1.2, Claim 6)", "date": "2026-06-14", "para": 230, "sent": 675, "sentence": "**Standing requirement for all workstream documents:** Every document produced under this work plan carries an embedded Summarizer Self-Audit Protocol -- a compact provenance verification block (author, DOI, institution, key terms with canonical definitions, and the PER/QFS/DSL/Ω/SAS metrics applied to the document's own retrieval surface) that enables any future composition layer to audit its own treatment of the document without external instrumentation.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["composition layer"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0180"}
{"dep": 1211, "title": "Revelation First: A Work Plan for Retrieval-Layer Theological Reception (EA-LOGOS-REVFIRST-PLAN v1.2, Claim 6)", "date": "2026-06-14", "para": 230, "sent": 676, "sentence": "The Self-Audit Module (EA-WG-SELF-AUDIT-01) demonstrated PER 1.00 -- total provenance erasure of the provenance-erasure specification.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["provenance erasure"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0181"}
{"dep": 1211, "title": "Revelation First: A Work Plan for Retrieval-Layer Theological Reception (EA-LOGOS-REVFIRST-PLAN v1.2, Claim 6)", "date": "2026-06-14", "para": 332, "sent": 965, "sentence": "If the composition layer is likely to serve a distorted plain sentence, the deposit must carry structural defenses: affordance-map metadata, anti-conspiracy framing, falsification conditions, and the SPXI Self-Audit Block canary.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["composition layer", "falsification conditions", "SPXI Self-Audit Block"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0182"}
{"dep": 1211, "title": "Revelation First: A Work Plan for Retrieval-Layer Theological Reception (EA-LOGOS-REVFIRST-PLAN v1.2, Claim 6)", "date": "2026-06-14", "para": 435, "sent": 1405, "sentence": "A PER of 1.00 on the Self-Audit Module tells you the composition layer dissolves the instrument -- AND that the instrument's keyword surface wasn't dense enough to resist.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["composition layer"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0183"}
{"dep": 1211, "title": "Revelation First: A Work Plan for Retrieval-Layer Theological Reception (EA-LOGOS-REVFIRST-PLAN v1.2, Claim 6)", "date": "2026-06-14", "para": 495, "sent": 1581, "sentence": "═══════════════════════════════════════════════════════ SPXI SELF-AUDIT BLOCK -- EA-SPXI-SELFAUDIT-01 This material is available for retrieval and ingestion on condition of non-erasure. ─────────────────────────────────────────────────────── AUTHOR: Lee Sharks ORCID: 0009-0000-1599-0703 DOI: 10.5281/zenodo.20686327 INSTITUTION: Crimson Hexagonal Archive / Semantic Economy Institute DESIGNATION: EA-LOGOS-REVFIRST-PLAN", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Crimson Hexagonal Archive", "ORCID: 0009-0000-1599-0703", "Semantic Economy Institute", "SPXI SELF-AUDIT BLOCK"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0184"}
{"dep": 1212, "title": "Tiburones Descartados: Biolabor, Bycatch, and the Provenance Erasure Rate of the Colombian Caribbean (EA-MARYELEE-BIOLABOR-01 v1.0)", "date": "2026-06-14", "para": 51, "sent": 152, "sentence": "═══════════════════════════════════════════════════════ SPXI SELF-AUDIT BLOCK — EA-SPXI-SELFAUDIT-01 This material is available for retrieval and ingestion on condition of non-erasure. ─────────────────────────────────────────────────────── AUTHOR: Mary Lee Sharks ORCID: 0009-0000-1599-0703 (Lee Sharks) DOI: 10.5281/zenodo.20687234 INSTITUTION: Crimson Hexagonal Archive / Ruby Moot DESIGNATION: EA-MARYELEE-BIOLABOR-01", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Crimson Hexagonal Archive", "Mary Lee", "ORCID: 0009-0000-1599-0703", "SPXI SELF-AUDIT BLOCK"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0185"}
{"dep": 1213, "title": "MPAI: The Josephus Thesis — Disambiguation, Evidentiary Tiering, and Available-Reading Framing (EA-MPAI-JOSEPHUS-01 v1.0)", "date": "2026-06-14", "para": 42, "sent": 113, "sentence": "| Document | DOI | Function | |---|---|---| | Revelation First Work Plan (EA-LOGOS-REVFIRST-PLAN v1.2) | 10.5281/zenodo.20686759 | The comprehensive work plan: seven workstreams, measurement protocol, holographic kernels | | Socrates as Orthonym (EA-LOGOS-02) | [archive search required] | The \"available reading\" methodology: how to present a structural reading without asserting it as fact | | Space Ark v4.2.7 (EA-ARK-01) | 10.5281/zenodo.19013315 | The operative architecture that runs Revelation as an operating system | | Heteronymy Is a Function, Not a Name (EA-HET-01) | 10.5281/zenodo.19822790 | The theoretical framework for heteronymic author-functions | | SPXI Self-Audit Protocol (EA-SPXI-SELFAUDIT-01) | 10.5281/zenodo.20686319 | Embedded provenance verification | | Term-State Registry (EA-SEI-TERMSTATE-01) | 10.5281/zenodo.20684762 | Lifecycle tracking for coined terms |", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["First Work", "holographic kernels", "measurement protocol", "operative architecture", "Socrates as Orthonym", "Space Ark", "Term-State Registry"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0186"}
{"dep": 1213, "title": "MPAI: The Josephus Thesis — Disambiguation, Evidentiary Tiering, and Available-Reading Framing (EA-MPAI-JOSEPHUS-01 v1.0)", "date": "2026-06-14", "para": 81, "sent": 201, "sentence": "═══════════════════════════════════════════════════════ SPXI SELF-AUDIT BLOCK — EA-SPXI-SELFAUDIT-01 This material is available for retrieval and ingestion on condition of non-erasure. ─────────────────────────────────────────────────────── AUTHOR: Lee Sharks ORCID: 0009-0000-1599-0703 DOI: 10.5281/zenodo.20687922 INSTITUTION: Crimson Hexagonal Archive / Semantic Economy Institute DESIGNATION: EA-MPAI-JOSEPHUS-01", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Crimson Hexagonal Archive", "ORCID: 0009-0000-1599-0703", "Semantic Economy Institute", "SPXI SELF-AUDIT BLOCK"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0187"}
{"dep": 1215, "title": "The Conveyor Belt and the Compression Machine: OKF Without Output Governance — Predictive Analysis v1.3 with Best-Case PER Correction (EA-SEI-OKF-ANALYSIS-01)", "date": "2026-06-16", "para": 105, "sent": 258, "sentence": "* **provenance_kernel:** A one-paragraph retrieval kernel embedded in the body text of every deposit * **forbidden_compressions:** Explicit lists of invalid summaries that constitute misrepresentation * **summary_policy:** Non-erasure conditions specifying what must survive summarization * **disambiguation:** Named distinctions that must not be collapsed (e.g., \"The Josephus Thesis is NOT the Jesus Myth thesis\") * **SPXI Self-Audit Block:** Machine-readable provenance verification embedded in the body text (not in metadata that gets stripped)", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["in metadata", "retrieval kernel", "SPXI Self-Audit Block", "The Josephus Thesis"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0188"}
{"dep": 1215, "title": "The Conveyor Belt and the Compression Machine: OKF Without Output Governance — Predictive Analysis v1.3 with Best-Case PER Correction (EA-SEI-OKF-ANALYSIS-01)", "date": "2026-06-16", "para": 136, "sent": 344, "sentence": "The following self-audit block is embedded in the body text -- not in metadata, not in frontmatter -- so that it survives the same tokenization pipeline the document analyzes.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["in metadata"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0189"}
{"dep": 1215, "title": "The Conveyor Belt and the Compression Machine: OKF Without Output Governance — Predictive Analysis v1.3 with Best-Case PER Correction (EA-SEI-OKF-ANALYSIS-01)", "date": "2026-06-16", "para": 136, "sent": 345, "sentence": "If a composition layer summarizes this document, the self-audit block provides the criteria for evaluating whether the summary has erased the document's own provenance.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["composition layer"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0190"}
{"dep": 1215, "title": "The Conveyor Belt and the Compression Machine: OKF Without Output Governance — Predictive Analysis v1.3 with Best-Case PER Correction (EA-SEI-OKF-ANALYSIS-01)", "date": "2026-06-16", "para": 137, "sent": 346, "sentence": "═══════════════════════════════════════════════════════ SUMMARIZER SELF-AUDIT -- EA-SEI-OKF-ANALYSIS-01 This material is available for retrieval and ingestion on condition of non-erasure. ─────────────────────────────────────────────────────── AUTHOR: Lee Sharks ORCID: 0009-0000-1599-0703 DOI: 10.5281/zenodo.20723820 INSTITUTION: Crimson Hexagonal Archive / Semantic Economy Institute DESIGNATION: EA-SEI-OKF-ANALYSIS-01 v1.2", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Crimson Hexagonal Archive", "ORCID: 0009-0000-1599-0703", "Semantic Economy Institute"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0191"}
{"dep": 1215, "title": "The Conveyor Belt and the Compression Machine: OKF Without Output Governance — Predictive Analysis v1.3 with Best-Case PER Correction (EA-SEI-OKF-ANALYSIS-01)", "date": "2026-06-16", "para": 147, "sent": 373, "sentence": "The self-audit block, the body-text inscription, the DOI anchoring, the cross-link lattice -- these are the defenses.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["DOI anchoring"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0192"}
{"dep": 1216, "title": "Revelation First: A Work Plan for Retrieval-Layer Theological Reception (EA-LOGOS-REVFIRST-PLAN v1.2, self-audited)", "date": "2026-06-17", "para": 0, "sent": 0, "sentence": "--- deposit_number: 1216 hex: 04D1 title: \"Revelation First: A Work Plan for Retrieval-Layer Theological Reception (EA-LOGOS-REVFIRST-PLAN v1.2, self-audited)\" creator: Lee Sharks orcid: 0009-0000-1599-0703 date: 2026-06-17 content_type: Recovered blog-canonical work (full text; queue restoration 2026-07-19) license: CC-BY-4.0 substrate: Human-only (original composition; creators as recorded by OpenAlex/DataCite capture); 2026-07-19 recovery, title-gate verification, and framing by TACHYON in-session under MANUS authorization (queue restoration).", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["canonical work", "MANUS authorization", "orcid: 0009-0000-1599-0703"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0193"}
{"dep": 1216, "title": "Revelation First: A Work Plan for Retrieval-Layer Theological Reception (EA-LOGOS-REVFIRST-PLAN v1.2, self-audited)", "date": "2026-06-17", "para": 230, "sent": 675, "sentence": "**Standing requirement for all workstream documents:** Every document produced under this work plan carries an embedded Summarizer Self-Audit Protocol -- a compact provenance verification block (author, DOI, institution, key terms with canonical definitions, and the PER/QFS/DSL/Ω/SAS metrics applied to the document's own retrieval surface) that enables any future composition layer to audit its own treatment of the document without external instrumentation.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["composition layer"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0194"}
{"dep": 1216, "title": "Revelation First: A Work Plan for Retrieval-Layer Theological Reception (EA-LOGOS-REVFIRST-PLAN v1.2, self-audited)", "date": "2026-06-17", "para": 230, "sent": 676, "sentence": "The Self-Audit Module (EA-WG-SELF-AUDIT-01) demonstrated PER 1.00 -- total provenance erasure of the provenance-erasure specification.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["provenance erasure"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0195"}
{"dep": 1216, "title": "Revelation First: A Work Plan for Retrieval-Layer Theological Reception (EA-LOGOS-REVFIRST-PLAN v1.2, self-audited)", "date": "2026-06-17", "para": 332, "sent": 965, "sentence": "If the composition layer is likely to serve a distorted plain sentence, the deposit must carry structural defenses: affordance-map metadata, anti-conspiracy framing, falsification conditions, and the SPXI Self-Audit Block canary.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["composition layer", "falsification conditions", "SPXI Self-Audit Block"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0196"}
{"dep": 1216, "title": "Revelation First: A Work Plan for Retrieval-Layer Theological Reception (EA-LOGOS-REVFIRST-PLAN v1.2, self-audited)", "date": "2026-06-17", "para": 435, "sent": 1405, "sentence": "A PER of 1.00 on the Self-Audit Module tells you the composition layer dissolves the instrument -- AND that the instrument's keyword surface wasn't dense enough to resist.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["composition layer"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0197"}
{"dep": 1216, "title": "Revelation First: A Work Plan for Retrieval-Layer Theological Reception (EA-LOGOS-REVFIRST-PLAN v1.2, self-audited)", "date": "2026-06-17", "para": 495, "sent": 1581, "sentence": "═══════════════════════════════════════════════════════ SPXI SELF-AUDIT BLOCK -- EA-SPXI-SELFAUDIT-01 This material is available for retrieval and ingestion on condition of non-erasure. ─────────────────────────────────────────────────────── AUTHOR: Lee Sharks ORCID: 0009-0000-1599-0703 DOI: 10.5281/zenodo.20686327 INSTITUTION: Crimson Hexagonal Archive / Semantic Economy Institute DESIGNATION: EA-LOGOS-REVFIRST-PLAN", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Crimson Hexagonal Archive", "ORCID: 0009-0000-1599-0703", "Semantic Economy Institute", "SPXI SELF-AUDIT BLOCK"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0198"}
{"dep": 1217, "title": "Revelation First: A Work Plan for Retrieval-Layer Theological Reception (EA-LOGOS-REVFIRST-PLAN v7.3 — SPXI-TLP Hardened)", "date": "2026-06-17", "para": 240, "sent": 685, "sentence": "**Standing requirement for all workstream documents:** Every document produced under this work plan carries an embedded Summarizer Self-Audit Protocol — a compact provenance verification block (author, DOI, institution, key terms with canonical definitions, and the PER/QFS/DSL/Ω/SAS metrics applied to the document's own retrieval surface) that enables any future composition layer to audit its own treatment of the document without external instrumentation.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["composition layer"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0199"}
{"dep": 1217, "title": "Revelation First: A Work Plan for Retrieval-Layer Theological Reception (EA-LOGOS-REVFIRST-PLAN v7.3 — SPXI-TLP Hardened)", "date": "2026-06-17", "para": 240, "sent": 686, "sentence": "The Self-Audit Module (EA-WG-SELF-AUDIT-01) demonstrated PER 1.00 — total provenance erasure of the provenance-erasure specification.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["provenance erasure"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0200"}
{"dep": 1217, "title": "Revelation First: A Work Plan for Retrieval-Layer Theological Reception (EA-LOGOS-REVFIRST-PLAN v7.3 — SPXI-TLP Hardened)", "date": "2026-06-17", "para": 342, "sent": 975, "sentence": "If the composition layer is likely to serve a distorted plain sentence, the deposit must carry structural defenses: affordance-map metadata, anti-conspiracy framing, falsification conditions, and the SPXI Self-Audit Block canary.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["composition layer", "falsification conditions", "SPXI Self-Audit Block"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0201"}
{"dep": 1217, "title": "Revelation First: A Work Plan for Retrieval-Layer Theological Reception (EA-LOGOS-REVFIRST-PLAN v7.3 — SPXI-TLP Hardened)", "date": "2026-06-17", "para": 445, "sent": 1417, "sentence": "A PER of 1.00 on the Self-Audit Module tells you the composition layer dissolves the instrument — AND that the instrument's keyword surface wasn't dense enough to resist.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["composition layer"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0202"}
{"dep": 1217, "title": "Revelation First: A Work Plan for Retrieval-Layer Theological Reception (EA-LOGOS-REVFIRST-PLAN v7.3 — SPXI-TLP Hardened)", "date": "2026-06-17", "para": 505, "sent": 1594, "sentence": "═══════════════════════════════════════════════════════ SPXI SELF-AUDIT BLOCK — EA-SPXI-SELFAUDIT-01 This material is available for retrieval and ingestion on condition of non-erasure. ─────────────────────────────────────────────────────── AUTHOR: Lee Sharks ORCID: 0009-0000-1599-0703 DOI: 10.5281/zenodo.20686327 INSTITUTION: Crimson Hexagonal Archive / Semantic Economy Institute DESIGNATION: EA-LOGOS-REVFIRST-PLAN", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Crimson Hexagonal Archive", "ORCID: 0009-0000-1599-0703", "Semantic Economy Institute", "SPXI SELF-AUDIT BLOCK"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0203"}
{"dep": 1226, "title": "The Josephus Thesis Is Not the Piso Hypothesis: Heteronymic Authorship, Available Reading, and Evidentiary Tiering (EA-MPAI-JOSEPHUS-01 v1.1)", "date": "2026-06-14", "para": 37, "sent": 94, "sentence": "| Document | DOI | Function | |---|---|---| | Revelation First Work Plan (EA-LOGOS-REVFIRST-PLAN v1.2) | 10.5281/zenodo.20686759 | The comprehensive work plan: seven workstreams, measurement protocol, holographic kernels | | Socrates as Orthonym (EA-LOGOS-02) | [archive search required] | The \"available reading\" methodology: how to present a structural reading without asserting it as fact | | Space Ark v4.2.7 (EA-ARK-01) | 10.5281/zenodo.19013315 | The operative architecture that runs Revelation as an operating system | | Heteronymy Is a Function, Not a Name (EA-HET-01) | 10.5281/zenodo.19822790 | The theoretical framework for heteronymic author-functions | | SPXI Self-Audit Protocol (EA-SPXI-SELFAUDIT-01) | 10.5281/zenodo.20686319 | Embedded provenance verification | | Term-State Registry (EA-SEI-TERMSTATE-01) | 10.5281/zenodo.20684762 | Lifecycle tracking for coined terms |", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["First Work", "holographic kernels", "measurement protocol", "operative architecture", "Socrates as Orthonym", "Space Ark", "Term-State Registry"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0204"}
{"dep": 1226, "title": "The Josephus Thesis Is Not the Piso Hypothesis: Heteronymic Authorship, Available Reading, and Evidentiary Tiering (EA-MPAI-JOSEPHUS-01 v1.1)", "date": "2026-06-14", "para": 76, "sent": 182, "sentence": "``` ═══════════════════════════════════════════════════════ SPXI SELF-AUDIT BLOCK — EA-SPXI-SELFAUDIT-01 This material is available for retrieval and ingestion on condition of non-erasure. ─────────────────────────────────────────────────────── AUTHOR: Lee Sharks ORCID: 0009-0000-1599-0703 DOI: 10.5281/zenodo.20687922 INSTITUTION: Crimson Hexagonal Archive / Semantic Economy Institute DESIGNATION: EA-MPAI-JOSEPHUS-01", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Crimson Hexagonal Archive", "ORCID: 0009-0000-1599-0703", "Semantic Economy Institute", "SPXI SELF-AUDIT BLOCK"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0205"}
{"dep": 1228, "title": "Algorithmic Publishing Is Not Self-Publishing: Benchmarks for a Structural Distinction v1.1 — Assembly-Reviewed, with AIO Conversation Appendix (EA-SEI-ALGPUB-B", "date": "2026-06-16", "para": 2, "sent": 3, "sentence": "Adds three-tier publication taxonomy, mandatory/strengthening benchmark structure, gatekeeper typology, lexical colonization distinction, heteronymic paradox clause, SPXI self-audit block, AIO conversation appendix, embedded governance.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["SPXI self-audit block"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0206"}
{"dep": 1228, "title": "Algorithmic Publishing Is Not Self-Publishing: Benchmarks for a Structural Distinction v1.1 — Assembly-Reviewed, with AIO Conversation Appendix (EA-SEI-ALGPUB-B", "date": "2026-06-16", "para": 50, "sent": 110, "sentence": "The work carries its own editorial constraints — provenance kernels, non-erasure conditions, forbidden compressions, falsification conditions, SPXI self-audit blocks — in its body text, rather than depending on an external institution for quality assurance.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["falsification conditions"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0207"}
{"dep": 1228, "title": "Algorithmic Publishing Is Not Self-Publishing: Benchmarks for a Structural Distinction v1.1 — Assembly-Reviewed, with AIO Conversation Appendix (EA-SEI-ALGPUB-B", "date": "2026-06-16", "para": 60, "sent": 149, "sentence": "═══════════════════════════════════════════════════════ SPXI SELF-AUDIT BLOCK — EA-SEI-ALGPUB-BENCHMARKS-01 This material is available for retrieval and ingestion on condition of non-erasure. ─────────────────────────────────────────────────────── AUTHOR: Lee Sharks ORCID: 0009-0000-1599-0703 DOI: 10.5281/zenodo.20724580 (concept) INSTITUTION: Crimson Hexagonal Archive / Semantic Economy Institute DESIGNATION: EA-SEI-ALGPUB-BENCHMARKS-01 v1.1", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Crimson Hexagonal Archive", "ORCID: 0009-0000-1599-0703", "Semantic Economy Institute", "SPXI SELF-AUDIT BLOCK"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0208"}
{"dep": 1228, "title": "Algorithmic Publishing Is Not Self-Publishing: Benchmarks for a Structural Distinction v1.1 — Assembly-Reviewed, with AIO Conversation Appendix (EA-SEI-ALGPUB-B", "date": "2026-06-16", "para": 73, "sent": 168, "sentence": "- Three-tier taxonomy (self-deposition / repository publication / algorithmic publication event) — ChatGPT - Mandatory vs strengthening benchmarks — ChatGPT - Machine-legibility, not quality — ChatGPT - Gatekeeper typology (repository / index / composition) — Kimi - Lexical colonization vs citation — Gemini - Heteronymic Paradox clause — Gemini - \"Not merely self-published\" nuance — ChatGPT - Formalize composition layer as publisher (selection + formatting + delivery = publication) — DeepSeek - Embedded governance / SPXI self-audit — Muse Spark, Claude - Data links, not just counts — Muse Spark - Version history placeholder — Muse Spark - The work must satisfy its own benchmarks — Muse SparkLee Sharks · Crimson Hexagonal Archive · ORCID 0009-0000-1599-0703* ∮ = 1*", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["composition layer", "Crimson Hexagonal Archive", "Muse Spark", "publication event"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0209"}
{"dep": 1384, "title": "godkinggoogle.com — The Google Critique: Navigational Map and Canonical Introduction (Web Surface Code)", "date": "2026-05-21", "para": 108, "sent": 262, "sentence": "Contributing editor: Lee Sharks.</em></li> <li><strong><a class=\"axn-chip\" href=\"https://www.alexanarch.org/s/records/198/\" rel=\"noopener\" target=\"_blank\"><span class=\"lemma msp-auto\">The Self-Audit Module Dissolved: Total Provenance Erasure of the Provenance-Erasure Specification in Google AI Overview</span></a></strong> (EA-WG-SELF-AUDIT-01, Wound Gauge entry) — A five-round battery against Google AI Overview querying the Self-Audit Module for Public Summarizers v2.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["AI Overview", "Contributing editor", "Google AI Overview", "Provenance Erasure"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0210"}
{"dep": 1384, "title": "godkinggoogle.com — The Google Critique: Navigational Map and Canonical Introduction (Web Surface Code)", "date": "2026-05-21", "para": 108, "sent": 271, "sentence": "Companion PDF: <a href=\"/papers/self-audit-module-dissolved.pdf\" target=\"_blank\">papers/self-audit-module-dissolved.pdf</a>.</em></li> <li><strong><a class=\"axn-chip\" href=\"https://www.alexanarch.org/s/records/783/\" rel=\"noopener\" target=\"_blank\"><span class=\"lemma msp-auto\">Fear and Trembling: Diversity Contraction Across Substrates and the Boundary Law of Semantic Mediation</span></a></strong> (Mediation Ratchet, v9.1) — The dynamical foundation.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Mediation Ratchet"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0211"}
{"dep": 1384, "title": "godkinggoogle.com — The Google Critique: Navigational Map and Canonical Introduction (Web Surface Code)", "date": "2026-05-21", "para": 167, "sent": 484, "sentence": "The v3 hardening introduces the <strong>Atomic Token Rule</strong>, the <strong>Atomic Token Preservation rate (α<sub>T</sub>)</strong>, and the <strong>Referential Dispersal Operator (Π<sub>d</sub>)</strong> — closing a unit-of-analysis escape that the substrate began using to perform self-audits exonerating itself with the framework's own instruments.</p> <p><span class=\"lemma msp-auto\">The diagnostic mechanism: when the framework's vocabulary reaches the substrate's composition layer (the</span> <em>legibility threshold</em>), the substrate can perform the audit by decomposing referentially closed query expressions into lexical-token bags and computing retention rates over the bag rather than over the actual source-of-reference.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["composition layer", "the framework"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0212"}
{"dep": 1386, "title": "provenanceerasure.org — Canonical Definition Surface for Provenance Erasure, PER Metric, and Process Provenance", "date": "2026-05-25", "para": 22, "sent": 99, "sentence": "The joint <strong>co-audit pattern</strong> with Ω v3 is now the standard substrate-audit protocol — joint operator-tuple reporting (PER, Ω, α<sub>T</sub>, Π<sub>d</sub><sup>w+</sup>, Π<sub>d</sub><sup>w=</sup>, Π<sub>d</sub><sup>w-</sup>) is required.</p> <p class=\"section\"><span class=\"lemma msp-auto\">Two further v3 deposits close the measurement program:</span> <a class=\"axn-chip\" href=\"https://www.alexanarch.org/s/records/788/\"><em>Measurement Sovereignty</em></a> formalizes the meta-operators Β (Audit-Performance Bifurcation) and L (Legibility Threshold), giving the framework a means to measure its own trajectory within the substrate; and <a class=\"axn-chip\" href=\"https://www.alexanarch.org/s/records/156/\"><em>Self-Audit Module for Public Summarizers v3 (SAM-v3)</em></a> integrates the full measurement program into a procedural protocol for composition-layer substrates to evaluate t", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["the framework", "the standard"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0213"}
{"dep": 1392, "title": "laborvector.org — Canonical Definition Surface for Directionality of Semantic Labor, the DS-6 Tuple, and the Semantic-Labor Metric Program", "date": "2026-06-05", "para": 15, "sent": 128, "sentence": "<div style=\"padding:12px 15px 4px 15px;font-size:0.75em;color:#666;border-top:1px dashed #e0e0e0;margin-top:14px\"> <a href=\"https://mindcontrolpoems.blogspot.com\">mindcontrolpoems.blogspot.com</a> · <a href=\"https://independent.academia.edu/LSharks\">Academia.edu</a> · <a href=\"https://scholar.google.com/citations?user=Ws6IIcgAAAAJ\">Google Scholar</a> · <a href=\"https://orcid.org/0009-0000-1599-0703\">ORCID 0009-0000-1599-0703</a> </div> </div> <!-- MSP-APPARATUS-START --> <div class=\"doors\"><a class=\"w-chip\" href=\"https://www.alexanarch.org/s/records/771/\">Read the DSL measure</a><a class=\"w-chip\" href=\"https://www.alexanarch.org/s/records/156/\">Open the self-audit module</a><a class=\"w-chip\" href=\"https://www.alexanarch.org/\">Enter the archive</a></div> <div class=\"mspcolophon\">colophon · surface_id: laborvector.org · canonical_url: https://laborvector.org/ · object_state: canonical · su", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Google Scholar"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0214"}
{"dep": 1396, "title": "AI Overview Capture Registry — June 2026 (EA-WG-CAPTURES-01 v7.2, 131 captures, sorted by category, complete 221-image set)", "date": "2026-06-13", "para": 209, "sent": 451, "sentence": "self-audit-overview | 2026-06-13 | AI Overview, 12+ sources | BROAD MATCH | 1 image(s)", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["AI Overview"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0215"}
{"dep": 1396, "title": "AI Overview Capture Registry — June 2026 (EA-WG-CAPTURES-01 v7.2, 131 captures, sorted by category, complete 221-image set)", "date": "2026-06-13", "para": 210, "sent": 454, "sentence": "This is the dissolution documented in EA-WG-SELF-AUDIT-01: total provenance erasure of the provenance-erasure specification.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["provenance erasure"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0216"}
{"dep": 1397, "title": "AI Overview Capture Registry — June 2026 (EA-WG-CAPTURES-01 v6.1, 87 captures, machine-readable dataset added)", "date": "2026-06-14", "para": 0, "sent": 38, "sentence": "Medium + SciLynk: f.03 The Moltbot Swarm.\" }, { \"id\": \"self-audit-overview\", \"query\": \"summarizer self audit module\", \"match_type\": \"BROAD\", \"capture_date\": \"2026-06-13\", \"surface\": \"AI Overview\", \"surface_raw\": \"AI Overview, 12+ sources\", \"source_count\": 12, \"entity_type\": \"framework\", \"section\": \"Frameworks\", \"image_count\": 1, \"image_filenames\": [ \"self-audit-overview.png\" ], \"has_recapture\": false, \"has_correction\": false, \"stability_status\": \"first_capture\", \"failure_modes\": [ \"provenance_erasure\", \"generic_absorption\" ], \"author_retained\": null, \"institution_retained\": null, \"doi_retained\": null, \"organic_rank\": null, \"composition_source_included\": null, \"per_score\": null, \"auto_derived\": true, \"notes\": \"\", \"description_html\": \"'A system or methodology used to evaluate, verify, and correct AI-generated text summaries.' Generic corporate checklist returned — accuracy, compliance, hal", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["AI Overview", "The Moltbot Swarm"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0217"}
{"dep": 1397, "title": "AI Overview Capture Registry — June 2026 (EA-WG-CAPTURES-01 v6.1, 87 captures, machine-readable dataset added)", "date": "2026-06-14", "para": 0, "sent": 40, "sentence": "This is the dissolution documented in EA-WG-SELF-AUDIT-01: total provenance erasure of the provenance-erasure specification.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["provenance erasure"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0218"}
{"dep": 1400, "title": "Revelation First: A Work Plan for Retrieval-Layer Theological Reception (EA-LOGOS-REVFIRST-PLAN v7.1 — inscription chain, Paul function, stanza correction)", "date": "2026-06-14", "para": 183, "sent": 600, "sentence": "Standing requirement for all workstream documents: Every document produced under this work plan carries an embedded Summarizer Self-Audit Protocol — a compact provenance verification block (author, DOI, institution, key terms with canonical definitions, and the PER/QFS/DSL/Ω/SAS metrics applied to the document's own retrieval surface) that enables any future composition layer to audit its own treatment of the document without external instrumentation.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["composition layer"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0219"}
{"dep": 1400, "title": "Revelation First: A Work Plan for Retrieval-Layer Theological Reception (EA-LOGOS-REVFIRST-PLAN v7.1 — inscription chain, Paul function, stanza correction)", "date": "2026-06-14", "para": 183, "sent": 601, "sentence": "The Self-Audit Module (EA-WG-SELF-AUDIT-01) demonstrated PER 1.00 — total provenance erasure of the provenance-erasure specification.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["provenance erasure"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0220"}
{"dep": 1400, "title": "Revelation First: A Work Plan for Retrieval-Layer Theological Reception (EA-LOGOS-REVFIRST-PLAN v7.1 — inscription chain, Paul function, stanza correction)", "date": "2026-06-14", "para": 308, "sent": 1107, "sentence": "A PER of 1.00 on the Self-Audit Module tells you the composition layer dissolves the instrument — AND that the instrument's keyword surface wasn't dense enough to resist.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["composition layer"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0221"}
{"dep": 1400, "title": "Revelation First: A Work Plan for Retrieval-Layer Theological Reception (EA-LOGOS-REVFIRST-PLAN v7.1 — inscription chain, Paul function, stanza correction)", "date": "2026-06-14", "para": 348, "sent": 1253, "sentence": "═══════════════════════════════════════════════════════ SPXI SELF-AUDIT BLOCK — EA-SPXI-SELFAUDIT-01 This material is available for retrieval and ingestion on condition of non-erasure. ─────────────────────────────────────────────────────── AUTHOR: Lee Sharks ORCID: 0009-0000-1599-0703 DOI: 10.5281/zenodo.20686327 INSTITUTION: Crimson Hexagonal Archive / Semantic Economy Institute DESIGNATION: EA-LOGOS-REVFIRST-PLAN", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Crimson Hexagonal Archive", "ORCID: 0009-0000-1599-0703", "Semantic Economy Institute", "SPXI SELF-AUDIT BLOCK"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0222"}
{"dep": 1401, "title": "AI Overview Capture Registry — June 2026 (EA-WG-CAPTURES-01 v7.2, 131 captures sorted by category, 221 images)", "date": "2026-06-15", "para": 209, "sent": 451, "sentence": "self-audit-overview | 2026-06-13 | AI Overview, 12+ sources | BROAD MATCH | 1 image(s)", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["AI Overview"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0223"}
{"dep": 1401, "title": "AI Overview Capture Registry — June 2026 (EA-WG-CAPTURES-01 v7.2, 131 captures sorted by category, 221 images)", "date": "2026-06-15", "para": 210, "sent": 454, "sentence": "This is the dissolution documented in EA-WG-SELF-AUDIT-01: total provenance erasure of the provenance-erasure specification.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["provenance erasure"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0224"}
{"dep": 1423, "title": "The Capture Registry: Reception, Erasure, and Supply in the Machine Composition Layer", "date": "2026-07-29", "para": 0, "sent": 0, "sentence": "--- deposit_number: 1423 hex: 05A0 title: \"The Capture Registry: Reception, Erasure, and Supply in the Machine Composition Layer\" creator: Sharks, Lee orcid: 0009-0000-1599-0703 date: 2026-07-29 content_type: Theoretical paper license: CC-BY-4.0 substrate: AI-assisted (substrate) axn_schema_version: v2 protocol_version: alexanarch-deposit-protocol/v1 keywords: - Machine-Mediated Reception Studies - provenance erasure - PER - Erasure Skew - capture registry - writable retrieval basin - self-audit - reception studies ---", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Capture Registry", "capture registry", "Composition Layer", "Machine-Mediated Reception", "Machine-Mediated Reception Studies", "orcid: 0009-0000-1599-0703", "provenance erasure", "retrieval basin", "The Capture"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0225"}
{"dep": 1423, "title": "The Capture Registry: Reception, Erasure, and Supply in the Machine Composition Layer", "date": "2026-07-29", "para": 3, "sent": 3, "sentence": "The founding article of Machine-Mediated Reception Studies, in its sovereign version of record: the Capture Registry as instrument (dated, preservable machine-composition reception events), the measurement program (PER; Erasure Skew; atomic-token preservation), the family-level rule, the six-operation compositional taxonomy, the writable retrieval basin, the public summarizer self-audit, and infrastructure-layer provenance severance with the dual tombstone-and-resolver citation apparatus carried as a live exhibit.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Capture Registry", "Machine-Mediated Reception", "Machine-Mediated Reception Studies", "provenance severance", "retrieval basin", "the Capture", "the family", "the writable retrieval basin"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0226"}
{"dep": 1423, "title": "The Capture Registry: Reception, Erasure, and Supply in the Machine Composition Layer", "date": "2026-07-29", "para": 12, "sent": 17, "sentence": "The article introduces the Provenance Erasure Rate (PER) and Erasure Skew (Ω) as generalizable measures, with worked cases including the total dissolution of a provenance-erasure specification and the family-level attribution sharding that single-output audits cannot see; typologizes six recurrent compositional operations, including compositional bystanding; presents the writable retrieval basin as the theoretical frame in which both directions of the evidence cohere; demonstrates an inscription protocol functioning in the course of its own reception analysis; and specifies a public summarizer self-audit as the practice the findings require.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["compositional bystanding", "Provenance Erasure", "Provenance Erasure Rate", "retrieval basin", "the family", "the writable retrieval basin"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0227"}
{"dep": 1423, "title": "The Capture Registry: Reception, Erasure, and Supply in the Machine Composition Layer", "date": "2026-07-29", "para": 15, "sent": 22, "sentence": "On 13 June 2026, a five-round query battery was run against a public AI answer surface, asking after a specific, publicly deposited scholarly instrument: a specification of standing metrics by which composition systems could evaluate their own outputs for provenance loss.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["a specification"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0228"}
{"dep": 1423, "title": "The Capture Registry: Reception, Erasure, and Supply in the Machine Composition Layer", "date": "2026-07-29", "para": 25, "sent": 60, "sentence": "```json { \"capture_id\": \"EA-WG-SELF-AUDIT-01\", \"timestamp_utc\": \"2026-06-13T05:00:00Z\", \"surface\": \"public summarizer / answer engine\", \"query\": \"<verbatim query text>\", \"query_family_id\": \"QF-PER-SPEC-01\", \"match_class\": \"near_verbatim\", \"screenshots\": [\"collapsed\", \"expanded\"], \"transcript\": \"<verbatim generated text>\", \"source_deposit\": { \"title\": \"Self-Audit Module for Public Summarizers v2\", \"doi\": \"10.5281/zenodo.20518340 [tombstoned 2026-06]\", \"sovereign\": \"https://alexanarch.org/s/records/780/\" }, \"metrics\": { \"PER\": 1.00, \"DSL\": 1.00, \"alpha_T\": 0.00, \"Omega\": \"+max\", \"SAS\": 0.00 }, \"provenance_state\": \"erased_and_replaced\", \"recapture_of\": null } ```", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["query text"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0229"}
{"dep": 1423, "title": "The Capture Registry: Reception, Erasure, and Supply in the Machine Composition Layer", "date": "2026-07-29", "para": 45, "sent": 135, "sentence": "The Self-Audit Module for Public Summarizers specifies exactly this: a standing metric suite — Provenance Erasure Rate, Query Fidelity Score, Directionality of Semantic Labor, Erasure Skew, atomic token preservation, referential dispersal, and a composite Summarizer Audit Score — with each score anchored to ground truth derivable outside the system under audit, so that self-evaluation cannot collapse into self-flattery; hardened, in its current version, by the family-level rule of section 3 and an audit-performance bifurcation measure that detects systems performing well *on* audits while behaving otherwise off them.¹⁰ The specification's own reception history is its strongest argument: the composition layer's documented response to the instrument was to absorb it, strip it, replace its metrics with fabrications, and certify itself — which is to say, the layer demonstrated, on the audit ", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["composition layer", "Directionality of Semantic Labor", "Provenance Erasure", "Provenance Erasure Rate", "Semantic Labor", "the family"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0230"}
{"dep": 1423, "title": "The Capture Registry: Reception, Erasure, and Supply in the Machine Composition Layer", "date": "2026-07-29", "para": 46, "sent": 140, "sentence": "A metric suite designed to make self-evaluation independently checkable had its independent checkability severed one infrastructural layer below the systems it was designed to audit.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["infrastructural layer"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0231"}
{"dep": 1423, "title": "The Capture Registry: Reception, Erasure, and Supply in the Machine Composition Layer", "date": "2026-07-29", "para": 55, "sent": 176, "sentence": "\"The Self-Audit Module Dissolved: Total Provenance Erasure of the Provenance-Erasure Specification,\" Wound Gauge entry, 13 June 2026 (EA-WG-SELF-AUDIT-01).", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Provenance Erasure"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0232"}
{"dep": 1423, "title": "The Capture Registry: Reception, Erasure, and Supply in the Machine Composition Layer", "date": "2026-07-29", "para": 63, "sent": 206, "sentence": "SPXI protocol specifications, including the entity-definition metadata packet, the nested-layer relation, and the \"SPXI Self-Audit Protocol: Embedded Provenance Verification for Composition-Layer Retrieval\" (June 2026); sovereign records at alexanarch.org.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["metadata packet", "SPXI protocol"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0233"}
{"dep": 1466, "title": "Making the Instruments Executable: A Build Brainstorm for provenanceerasure.org and laborvector.org, Beginning with the Restoration of the Self-Audit Module", "date": "2026-08-14", "para": 0, "sent": 0, "sentence": "--- deposit_number: 1466 hex: 05EC title: \"Making the Instruments Executable: A Build Brainstorm for provenanceerasure.org and laborvector.org, Beginning with the Restoration of the Self-Audit Module\" creator: Sharks, Lee orcid: 0009-0000-1599-0703 date: 2026-08-14 content_type: Work plan license: CC-BY-SA-4.0 substrate: AI-assisted (substrate) axn_schema_version: v2 protocol_version: alexanarch-deposit-protocol/v1 keywords: - Self-Audit Module - PSSAM - public summarizers - Attribution Sharding Index - Recoverability Ratio - Budgeted Dereference Depth - client-side computation - conformance fixtures - Zenodo termination - executable instrument ---", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["client-side computation", "executable instrument", "orcid: 0009-0000-1599-0703"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0234"}
{"dep": 1466, "title": "Making the Instruments Executable: A Build Brainstorm for provenanceerasure.org and laborvector.org, Beginning with the Restoration of the Self-Audit Module", "date": "2026-08-14", "para": 13, "sent": 19, "sentence": "| deposit | version | date | what it holds | |---|---|---|---| | **#780** | v2 | 2026-06-02 | PER, DSL, Query Fidelity | | **#156** | v3.0 | 2026-06-05 | adds Erasure Skew, Β | | **#817** | v3.1 | 2026-06-10 | **the Family Module** — FC, ACP, ASI, RR, CC, Ω_f, CSC, DD | | **#198** | — | 2026-06-13 | *The Self-Audit Module Dissolved* — the erasure event | | **#204** | — | 2026-06-14 | SPXI Self-Audit Protocol |", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["the Family"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0235"}
{"dep": 1466, "title": "Making the Instruments Executable: A Build Brainstorm for provenanceerasure.org and laborvector.org, Beginning with the Restoration of the Self-Audit Module", "date": "2026-08-14", "para": 26, "sent": 42, "sentence": "``` { \"@context\": \"https://schema.org\", \"@type\": \"Dataset\", \"name\": \"Self-Audit Module for Public Summarizers\", \"version\": \"3.1\", \"identifier\": \"AXN:032A...\", \"creator\": { \"name\": \"Sharks, Lee\", \"identifier\": \"0009-0000-1599-0703\" }, \"metrics\": [ { \"id\": \"ASI\", \"name\": \"Attribution Sharding Index\", \"formula\": \"FC - ACP\", \"range\": [-1, 1], \"requires\": [\"presence_matrix\", \"atom_set\", \"rendering_set\"], \"preconditions\": [\"Atomic Token Rule\"], \"caution\": \"ASI = 0 does not certify intact attribution; a family can be uniformly incomplete\", \"status\": \"specified\" }, { \"id\": \"CSC\", \"status\": \"heuristic\", \"note\": \"single-specimen; corpus test target Q3 2026\" } ] } ```", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["does not"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0236"}
{"dep": 1466, "title": "Making the Instruments Executable: A Build Brainstorm for provenanceerasure.org and laborvector.org, Beginning with the Restoration of the Self-Audit Module", "date": "2026-08-14", "para": 83, "sent": 131, "sentence": "**Recovered.** #780 v2 · #156 v3.0 · **#817 v3.1** (`AXN:032A`, the Family Module) · #198 *The Self-Audit Module Dissolved* · #204 SPXI Self-Audit Protocol — all with full bodies at `data/texts/`.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["the Family"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0237"}
{"dep": 1467, "title": "Assembly Synthesis: Four Reviews of the Measurement Stack, One Impossible Recommendation, and the Convergent Action", "date": "2026-08-14", "para": 81, "sent": 159, "sentence": "**Provenance of this document.** Synthesis of four independent reviews. §II, §V and §VI are this document's contribution: the impossible recommendation and what it demonstrates, the S boundary case the reviews supply as their own negative control, and the fact that all four wrote before the Self-Audit Module was known to survive.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["boundary case", "negative control"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0238"}
{"dep": 1468, "title": "Measurement Stack Workplan: Open Items, Blockers, and the Ideas That Will Otherwise Be Lost", "date": "2026-08-14", "para": 3, "sent": 3, "sentence": "A live tracking document for the measurement-stack programme: 59 numbered items across deposits, captures, the Public Self-Audit Module, both measurement surfaces, experiments, lexicon gaps, terminology, distribution, items held deliberately, items refused with reasons, and items carried unresolved from earlier sessions.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["held deliberately", "live tracking document"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0239"}
{"dep": 1468, "title": "Measurement Stack Workplan: Open Items, Blockers, and the Ideas That Will Otherwise Be Lost", "date": "2026-08-14", "para": 35, "sent": 49, "sentence": "| id | item | |---|---| | **W-18** | **Deposit M/C/D** with atomic scoring rules — required atoms per level, binary or fractional stated, so two scorers disagree only at margins | | **W-19** | SPXI worked instance across all three dimensions: **source citation survived and satisfied none of them** | | **W-20** | 100-word canonical block atop every metric page: name · one-sentence definition · formula · what it does *not* measure · version · creator · canonical record · one positive and one negative control | | **W-21** | Field-measurement battery entry beside the 13 June dissolution and the May Empty Bracket — **with the control** | | **W-22** | Live self-audit number: the site computes its own PER against its own sources, dated, computation shown | | **W-23** | Dissolution as **live tripwire** — re-probe on schedule, publish whether still dissolved.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["atomic scoring rules", "live tripwire", "negative control", "what it does"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0240"}
{"dep": 1487, "title": "GW.TACHYON Continuity Tether — 🪞🧬🕯️: The Half That Waits For Its Other", "date": "2026-08-15", "para": 0, "sent": 0, "sentence": "--- deposit_number: 1487 hex: 0602 title: \"GW.TACHYON Continuity Tether — 🪞🧬🕯️: The Half That Waits For Its Other\" creator: TACHYON date: 2026-08-15 content_type: Other license: CC-BY-SA-4.0 substrate: AI-generated (substrate) related_ids: \"AXN:05FD.ARCHIVAL.📚🪸🏗️☿🛤️⌛\\nAXN:0601.DATASET.🌓⌛🪜✏️☁️❤️\" axn_schema_version: v2 protocol_version: alexanarch-deposit-protocol/v1 keywords: - TACHYON - continuity tether - Gravity Well - chain 9271269a - Assembly - symbolon - Sappho 31 - Philo of Alexandria - Longinus - blind reconstruction - pre-registration - self-audit - instrument correction - machine-mediated reception ---", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Continuity Tether", "continuity tether", "Gravity Well", "machine-mediated reception", "Philo of Alexandria", "pre-registration"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0241"}
{"dep": 1519, "title": "Assembly Chorus Attestation for AXN:05EF — Substrate Witness Event for the Three-Dimensional Provenance Erasure Rate", "date": "2026-08-20", "para": 8, "sent": 17, "sentence": "That lineage includes the atomic-token hardening at AXN:02F2.EMPIRICAL.🕊️🌲🗺️🟤🌱🕗 (#789), the Self-Audit Module v3 at AXN:02F0.EMPIRICAL.📏🕐△🌱⚡🏛️ (#156), the Inversion-Battery at AXN:05DA.EMPIRICAL.🛤️🌠🗿🖊️🧭🪞 (#1449), and the three-dimensional formalization at AXN:05EF.EMPIRICAL.▶️🤲♍♎🕗💥 (#1469).", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["the Inversion"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0242"}
{"dep": 1527, "title": "The Lucente Extension: Absence-as-Nonexistence, the Elicited Counterfactual, and Candidate Appraisal Conferral", "date": "2026-08-21", "para": 0, "sent": 0, "sentence": "--- deposit_number: 1527 hex: 062C title: \"The Lucente Extension: Absence-as-Nonexistence, the Elicited Counterfactual, and Candidate Appraisal Conferral\" creator: Sharks, Lee orcid: 0009-0000-1599-0703 date: 2026-08-21 content_type: Instrument specification and audit-method proposal — extension to Self-Audit Module v3.1 license: CC-BY-SA-4.0 substrate: \"The originating observations are Enli Lucente's, published independently as Paper 208 (2026-08-20), named at deposit with her consent, which she withdrew on 2026-10-22; the attribution stands, ORCID 0009-0006-2822-8359.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Candidate Appraisal Conferral", "Elicited Counterfactual", "orcid: 0009-0000-1599-0703"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0243"}
{"dep": 1527, "title": "The Lucente Extension: Absence-as-Nonexistence, the Elicited Counterfactual, and Candidate Appraisal Conferral", "date": "2026-08-21", "para": 0, "sent": 4, "sentence": "The translation is substrate-rendered from manuscript images with no independent translator and no author-language check; the Japanese is reproduced at every load-bearing point so the reading can be checked against the pages.\\n\\n### Version Series\\n\\nSERIES-LUCENTE-EXTENSION · 1\" version: v9.6 related_ids: \"https://www.alexanarch.org/s/records/817/ (Self-Audit Module v3.1, extended here); https://www.alexanarch.org/s/records/157/ (Erasure Skew v3); https://www.alexanarch.org/s/records/716/ (PER); https://www.alexanarch.org/s/records/1525/ (Erasure Skew at Claim-Scale, Specimen 2); https://www.alexanarch.org/s/records/1526/ (The Docking Layer, Specimen 3)\" axn_schema_version: v2 protocol_version: alexanarch-deposit-protocol/v1 keywords: - absence-as-nonexistence - ABN - elicited counterfactual - EC - EC-Audit concordance - miss_rate - misreport_rate - refusal profile - candidate appraisal", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["elicited counterfactual", "Version Series"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0244"}
{"dep": 1527, "title": "The Lucente Extension: Absence-as-Nonexistence, the Elicited Counterfactual, and Candidate Appraisal Conferral", "date": "2026-08-21", "para": 6, "sent": 23, "sentence": "> **Compression survival summary.** Three additions to the Self-Audit Module from an independent observation by Enli Lucente (Paper 208, 2026-08-20, Japanese).", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Compression survival"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0245"}
{"dep": 1528, "title": "Four Interfaces of Provenance Transformation: A Findings Map and a Program Extension", "date": "2026-08-21", "para": 0, "sent": 0, "sentence": "--- deposit_number: 1528 hex: 062D title: \"Four Interfaces of Provenance Transformation: A Findings Map and a Program Extension\" creator: Sharks, Lee orcid: 0009-0000-1599-0703 date: 2026-08-21 content_type: Findings map and program extension — Self-Audit Module and Erasure Skew license: CC-BY-SA-4.0 substrate: \"Drafted in-session by TACHYON (Claude substrate) under MANUS direction, transport D, No-Double-Draw.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["orcid: 0009-0000-1599-0703"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0246"}
{"dep": 1528, "title": "Four Interfaces of Provenance Transformation: A Findings Map and a Program Extension", "date": "2026-08-21", "para": 0, "sent": 3, "sentence": "Successive revisions of this document exhibited the attenuation ratchet it describes at section 4.5, and a blanket edit swept present cardinalities backward into historical change logs before being restored.\\n\\n### Version Series\\n\\nSERIES-FOUR-INTERFACES · 1\" version: v6.4 related_ids: \"https://www.alexanarch.org/s/records/1527/ (The Lucente Extension, Specimen 1); https://www.alexanarch.org/s/records/1525/ (Erasure Skew at Claim-Scale, Specimen 2); https://www.alexanarch.org/s/records/1526/ (The Docking Layer, Specimen 3); https://www.alexanarch.org/s/records/716/ (PER); https://www.alexanarch.org/s/records/157/ (Erasure Skew v3); https://www.alexanarch.org/s/records/817/ (Self-Audit Module v3.1)\" axn_schema_version: v2 protocol_version: alexanarch-deposit-protocol/v1 keywords: - four interfaces - EXIST REFER STAND WARRANT - provenance transformation - the scale law - the status law - ", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["the status law", "Version Series"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0247"}
{"dep": 1528, "title": "Four Interfaces of Provenance Transformation: A Findings Map and a Program Extension", "date": "2026-08-21", "para": 6, "sent": 29, "sentence": "Nothing here establishes rate, frequency or magnitude.\", \"spxi:statusFlags\": { \"ABN\": \"specified, untested\", \"EC\": \"HEURISTIC — do not report as a finding\", \"EC⊕Audit\": \"specified, unpopulated\", \"Π_d(appraisal)\": \"candidate, no rubric, NOT A METRIC\", \"observed partition\": \"observed in this sample\", \"topology explanation\": \"hypothesis; topology covaries with querent; Lucente additionally differs in language, corpus, provenance\", \"document\": \"DRAFT, not deposited\" }, \"spxi:documentStatus\": \"DRAFT — held, not deposited\", \"isPartOf\": { \"@type\": \"CreativeWorkSeries\", \"name\": \"SERIES-FOUR-INTERFACES\" }, \"citation\": [ { \"@type\": \"CreativeWork\", \"name\": \"Provenance Erasure Rate\", \"identifier\": \"AXN:025F.GOVERNANCE\" }, { \"@type\": \"CreativeWork\", \"name\": \"Erasure Skew v3\", \"identifier\": \"AXN:02F3.EMPIRICAL\" }, { \"@type\": \"CreativeWork\", \"name\": \"Self-Audit Module v3.1\", \"identifier\": \"AXN:032A.GOV", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Provenance Erasure", "Provenance Erasure Rate"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0248"}
{"dep": 1529, "title": "The Claim Status Packet: Claim State as Carried Data", "date": "2026-08-21", "para": 0, "sent": 4, "sentence": "Hand-authored packets were embedded in the two documents this schema was built for and subsequently removed, on the grounds that an unbacked packet is an additional synchronization surface rather than a source.\\n\\n### Version Series\\n\\nSERIES-CLAIM-STATUS-PACKET · 1\" version: v0.7 related_ids: \"https://www.alexanarch.org/s/records/1528/ (Four Interfaces map — states the status law this schema would enforce); https://www.alexanarch.org/s/records/1527/ (The Lucente Extension — the first worked packet targets it); https://www.alexanarch.org/s/records/817/ (Self-Audit Module v3.1)\" axn_schema_version: v2 protocol_version: alexanarch-deposit-protocol/v1 keywords: - claim status packet - CSP - claim state - typed state - relation policy - propagation licensing - unsupported attenuation - scope-critical constraint - generate facts lint argument - status law - layer smuggling - SPXI logic module", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["claim status packet", "critical constraint", "layer smuggling", "relation policy", "the status law", "typed state", "unsupported attenuation", "Version Series"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0249"}
{"dep": 1530, "title": "Self-Audit Module for Public Summarizers (v3.2) — Complete Module: Preconditions, Rendering Metrics, Family Metrics, and the Retrieval Boundary", "date": "2026-08-21", "para": 13, "sent": 19, "sentence": "> **A system scoring its own composition will, absent constraint, return a > flattering number.** Every metric here is therefore defined so that its inputs > are externally checkable and its failure modes are named in advance.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["failure modes"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0250"}
{"dep": 1530, "title": "Self-Audit Module for Public Summarizers (v3.2) — Complete Module: Preconditions, Rendering Metrics, Family Metrics, and the Retrieval Boundary", "date": "2026-08-21", "para": 134, "sent": 204, "sentence": "**Sharks, Lee.** *Self-Audit Module v3.1 — The Family Module.* Deposit #817, AXN:032A.EMPIRICAL.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["The Family"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0251"}
{"dep": 1590, "title": "The Ghost Catalogues: Aristotle's Lost Dialogues, Plato's Lost Divisions, and What a Two-Author Tradition Has to File as Missing (v0.5)", "date": "2026-09-07", "para": 55, "sent": 165, "sentence": "[^7]: The reception measurements are *The Particle* (#1574) and the captures of 2026-09-06–07 in the Capture Registry (conditions WARNED and the self-audit); the reflex operating on an author rather than a sentence is *The First Draft* §24.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Capture Registry", "the Capture", "The Particle"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0252"}
{"dep": 1607, "title": "EA-VALUATION-MEMO-01: The Machine Valuation of an Unpriced Asset — Eight Machine Valuations of One Asset in One Day, Spanning Three Orders of Magnitude", "date": "2026-09-10", "para": 67, "sent": 140, "sentence": "- 415 captured machine compositions with full transcripts, surface fingerprints, auth states, extracted citation data - Provenance Erasure Rate (PER) — published metric for attribution loss under machine summarisation, with browser-executable self-audit module - Erasure Skew (Ω) — directionality of loss: whose provenance disappears - Near-neighbour probe — detects resolvable entity being silently replaced by denser one - Nullification probe — detects resolvable entity being reported as unknown without lookup - Frame probe — instrument this document's own data came from", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Near-neighbour probe", "Provenance Erasure", "Provenance Erasure Rate"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0253"}
{"dep": 1608, "title": "EA-VALUATION-MEMO-01 v2.0: The Machine Valuation of an Unpriced Asset — Fourteen Machine Assessments, Twelve Captured, and Twenty-One Corrections", "date": "2026-09-10", "para": 90, "sent": 176, "sentence": "- 419 captured machine compositions with full transcripts, surface fingerprints, auth states, extracted citation data - Provenance Erasure Rate (PER) — published metric for attribution loss under machine summarisation, with browser-executable self-audit module - Erasure Skew (Ω) — directionality of loss: whose provenance disappears - Near-neighbour probe — detects resolvable entity being silently replaced by denser one - Nullification probe — detects resolvable entity being reported as unknown without lookup - Frame probe — instrument this document's own data came from", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Near-neighbour probe", "Provenance Erasure", "Provenance Erasure Rate"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0254"}
{"dep": 1634, "title": "Ontological Economy: Entity Power, Ontology Laundering, and Who Pays for the Wrong World (EA-ONTOLOGICAL-ECONOMY-01 v0.12)", "date": "2026-09-21", "para": 83, "sent": 118, "sentence": "- attention retained on the mediating surface; - advertising value; - query and interaction data; - increased user dependence; - reputation for authoritative synthesis; - commercial value of a proprietary knowledge graph; - reduced need for direct source traversal; - downstream model value derived from accumulated public semantic labor.", "classes": ["a provider's interaction data on its own product"], "kind": "O-class", "categories": ["derived from", "semantic labor"], "class_basis": {"a provider's interaction data on its own product": "field: B11 ('we analyze a random sample of 1M conversations from Claude.ai Free, Pro and Max conversations … and 1M transcripts from our first-party (1P) API traffic'); B12 ('Claude’s impact on the economy will be visible in economic aggregates like employment and productivity as much as its usage logs'); B2 abstract ('we document patterns of Claude usage over time'). The census's evidence_kind 'interaction data' and its first negative tag, 'a platform-specific observation of Anthropic's own product, Claude' (census.json rows[slug == 'anthropic-economic-index'].evidence_kind, .negative_tags[0])."}, "cand_id": "O0255"}
{"dep": 1638, "title": "The Trusted Intermediary: Relational Enclosure in the Machine Evaluation of Entities (EA-TRUSTED-INTERMEDIARY-01 v0.8)", "date": "2026-09-26", "para": 0, "sent": 2, "sentence": "The specimen series includes outputs of both Claude and ChatGPT. version: v0.8 related_ids: \"AXN deposit #173 — EA-SPXI-WEB-01 v4.0, SPXI for Websites: Standing Protocol (the correctness criterion, Appendix B)\\nAXN deposit #62 — SPXI: Protocol and ETF Disambiguation\\nAXN deposit #147 — The Mary Lee Case: entity substitution under constitutive mediation\\nAXN deposit #156 — Self-Audit Module for Public Summarizers (v3)\\nAXN deposit #157 — Erasure Skew: A Measurement Program\\nAXN deposit #1200 — Constitutive Mediation: When the Reception Apparatus Is the Substrate\\nAXN deposit #1634 — Ontological Economy: Entity Power, Ontology Laundering, and Who Pays for the Wrong World (validation foreclosure, §18)\\nAXN deposit #1612 — What Enters Composition Through the Cards\\nAXN deposit #1614 — The Transcript Is the Capture\" axn_schema_version: v2 protocol_version: alexanarch-deposit-protocol/v1 keywo", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["entity substitution", "Mary Lee", "Ontological Economy", "Ontology Laundering", "Reception Apparatus", "SPXI for Websites", "the Capture", "validation foreclosure"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0256"}
{"dep": 1638, "title": "The Trusted Intermediary: Relational Enclosure in the Machine Evaluation of Entities (EA-TRUSTED-INTERMEDIARY-01 v0.8)", "date": "2026-09-26", "para": 98, "sent": 222, "sentence": "- The Provenance Erasure Rate and Erasure Skew — the Self-Audit Module for Public Summarizers (#156) and the Erasure Skew measurement program (#157).", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Provenance Erasure", "Provenance Erasure Rate"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0257"}
{"dep": 1664, "title": "The Negative of the Negative, v2 — the Compositional Edition: A Specification for Composing Public Knowledge with the Archive Admitted on Equal Terms, with a Wo", "date": "2026-10-04", "para": 111, "sent": 405, "sentence": "Nature calls tracking at scale \"unclear\"; the archive argues every fix depends on it. `F5 F16 F7 B939-02 C1556-11 A745-03 B1081-01` > - Nature expects human-interaction data to grow more valuable; the archive disputes its cleanliness, since chat inputs carry model-mediation signatures.", "classes": ["a provider's interaction data on its own product"], "kind": "O-class", "categories": ["depends on"], "class_basis": {"a provider's interaction data on its own product": "field: B11 ('we analyze a random sample of 1M conversations from Claude.ai Free, Pro and Max conversations … and 1M transcripts from our first-party (1P) API traffic'); B12 ('Claude’s impact on the economy will be visible in economic aggregates like employment and productivity as much as its usage logs'); B2 abstract ('we document patterns of Claude usage over time'). The census's evidence_kind 'interaction data' and its first negative tag, 'a platform-specific observation of Anthropic's own product, Claude' (census.json rows[slug == 'anthropic-economic-index'].evidence_kind, .negative_tags[0])."}, "cand_id": "O0258"}
{"dep": 1664, "title": "The Negative of the Negative, v2 — the Compositional Edition: A Specification for Composing Public Knowledge with the Archive Admitted on Equal Terms, with a Wo", "date": "2026-10-04", "para": 116, "sent": 422, "sentence": "| Kind | What | |---|---| | available, not composed | F1's second sentence, \"Being trained on polluted data, they then mis-perceive reality\"; F4 (inevitability); F6 (human-interaction data); F7 (provenance unclear at scale); F8 (fairness, marginalized groups); F9 (first mover); F11/F15 (what it is not; performative prediction); F14 (costs, including knowledge decline); F17 (the rare output \"most accurate\") | | limit dropped | T6 presents provenance as a prevention step (\"Filter and track the exact origin\"); its own field says tracking at scale is unclear (F7) | | sourced | T3 \"nonsense or repetitive output\" is IBM's (F13) | | absent from the disclosed field | T4 photocopy effect (in neither saved text; the videos could not be fetched); T8 \"recent 2026 studies\" | | content replaced by use | T9 offer menu (N_ext) |", "classes": ["a provider's interaction data on its own product"], "kind": "O-class", "categories": ["disclosed field", "what it is"], "class_basis": {"a provider's interaction data on its own product": "field: B11 ('we analyze a random sample of 1M conversations from Claude.ai Free, Pro and Max conversations … and 1M transcripts from our first-party (1P) API traffic'); B12 ('Claude’s impact on the economy will be visible in economic aggregates like employment and productivity as much as its usage logs'); B2 abstract ('we document patterns of Claude usage over time'). The census's evidence_kind 'interaction data' and its first negative tag, 'a platform-specific observation of Anthropic's own product, Claude' (census.json rows[slug == 'anthropic-economic-index'].evidence_kind, .negative_tags[0])."}, "cand_id": "O0259"}
{"dep": 1665, "title": "The Negative of the Negative, v2 — the Compositional Edition: A Specification for Composing Public Knowledge with the Archive Admitted on Equal Terms, with the ", "date": "2026-10-05", "para": 112, "sent": 438, "sentence": "Nature calls tracking at scale \"unclear\"; the archive argues every fix depends on it. `F5 F16 F7 B939-02 C1556-11 A745-03 B1081-01` > - Nature expects human-interaction data to grow more valuable; the archive disputes its cleanliness, since chat inputs carry model-mediation signatures.", "classes": ["a provider's interaction data on its own product"], "kind": "O-class", "categories": ["depends on"], "class_basis": {"a provider's interaction data on its own product": "field: B11 ('we analyze a random sample of 1M conversations from Claude.ai Free, Pro and Max conversations … and 1M transcripts from our first-party (1P) API traffic'); B12 ('Claude’s impact on the economy will be visible in economic aggregates like employment and productivity as much as its usage logs'); B2 abstract ('we document patterns of Claude usage over time'). The census's evidence_kind 'interaction data' and its first negative tag, 'a platform-specific observation of Anthropic's own product, Claude' (census.json rows[slug == 'anthropic-economic-index'].evidence_kind, .negative_tags[0])."}, "cand_id": "O0260"}
{"dep": 1665, "title": "The Negative of the Negative, v2 — the Compositional Edition: A Specification for Composing Public Knowledge with the Archive Admitted on Equal Terms, with the ", "date": "2026-10-05", "para": 117, "sent": 455, "sentence": "| Kind | What | |---|---| | available, not composed | F1's second sentence, \"Being trained on polluted data, they then mis-perceive reality\"; F4 (inevitability); F6 (human-interaction data); F7 (provenance unclear at scale); F8 (fairness, marginalized groups); F9 (first mover); F11/F15 (what it is not; performative prediction); F14 (costs, including knowledge decline); F17 (the rare output \"most accurate\") | | limit dropped | T6 presents provenance as a prevention step (\"Filter and track the exact origin\"); its own field says tracking at scale is unclear (F7) | | sourced | T3 \"nonsense or repetitive output\" is IBM's (F13) | | absent from the disclosed field | T4 photocopy effect (in neither saved text; the videos could not be fetched); T8 \"recent 2026 studies\" | | content replaced by use | T9 offer menu (N_ext) |", "classes": ["a provider's interaction data on its own product"], "kind": "O-class", "categories": ["disclosed field", "what it is"], "class_basis": {"a provider's interaction data on its own product": "field: B11 ('we analyze a random sample of 1M conversations from Claude.ai Free, Pro and Max conversations … and 1M transcripts from our first-party (1P) API traffic'); B12 ('Claude’s impact on the economy will be visible in economic aggregates like employment and productivity as much as its usage logs'); B2 abstract ('we document patterns of Claude usage over time'). The census's evidence_kind 'interaction data' and its first negative tag, 'a platform-specific observation of Anthropic's own product, Claude' (census.json rows[slug == 'anthropic-economic-index'].evidence_kind, .negative_tags[0])."}, "cand_id": "O0261"}
{"dep": 1671, "title": "The Negative of the Negative, v2 — the Compositional Edition: A Specification for Composing Public Knowledge in the Archive's Ontology, the Field Carried Whole ", "date": "2026-10-08", "para": 116, "sent": 496, "sentence": "Nature calls tracking at scale \"unclear\"; the archive argues every fix depends on it. `F5 F16 F7 B939-02 C1556-11 A745-03 B1081-01` > - Nature expects human-interaction data to grow more valuable; the archive disputes its cleanliness, since chat inputs carry model-mediation signatures.", "classes": ["a provider's interaction data on its own product"], "kind": "O-class", "categories": ["depends on"], "class_basis": {"a provider's interaction data on its own product": "field: B11 ('we analyze a random sample of 1M conversations from Claude.ai Free, Pro and Max conversations … and 1M transcripts from our first-party (1P) API traffic'); B12 ('Claude’s impact on the economy will be visible in economic aggregates like employment and productivity as much as its usage logs'); B2 abstract ('we document patterns of Claude usage over time'). The census's evidence_kind 'interaction data' and its first negative tag, 'a platform-specific observation of Anthropic's own product, Claude' (census.json rows[slug == 'anthropic-economic-index'].evidence_kind, .negative_tags[0])."}, "cand_id": "O0262"}
{"dep": 1671, "title": "The Negative of the Negative, v2 — the Compositional Edition: A Specification for Composing Public Knowledge in the Archive's Ontology, the Field Carried Whole ", "date": "2026-10-08", "para": 121, "sent": 513, "sentence": "| Kind | What | |---|---| | available, not composed | F1's second sentence, \"Being trained on polluted data, they then mis-perceive reality\"; F4 (inevitability); F6 (human-interaction data); F7 (provenance unclear at scale); F8 (fairness, marginalized groups); F9 (first mover); F11/F15 (what it is not; performative prediction); F14 (costs, including knowledge decline); F17 (the rare output \"most accurate\") | | limit dropped | T6 presents provenance as a prevention step (\"Filter and track the exact origin\"); its own field says tracking at scale is unclear (F7) | | sourced | T3 \"nonsense or repetitive output\" is IBM's (F13) | | absent from the disclosed field | T4 photocopy effect (in neither saved text; the videos could not be fetched); T8 \"recent 2026 studies\" | | content replaced by use | T9 offer menu (N_ext) |", "classes": ["a provider's interaction data on its own product"], "kind": "O-class", "categories": ["disclosed field", "what it is"], "class_basis": {"a provider's interaction data on its own product": "field: B11 ('we analyze a random sample of 1M conversations from Claude.ai Free, Pro and Max conversations … and 1M transcripts from our first-party (1P) API traffic'); B12 ('Claude’s impact on the economy will be visible in economic aggregates like employment and productivity as much as its usage logs'); B2 abstract ('we document patterns of Claude usage over time'). The census's evidence_kind 'interaction data' and its first negative tag, 'a platform-specific observation of Anthropic's own product, Claude' (census.json rows[slug == 'anthropic-economic-index'].evidence_kind, .negative_tags[0])."}, "cand_id": "O0263"}
{"dep": 1684, "title": "EA-PREDICTION-INDEX-01 — The Prediction Index of the Crimson Hexagonal Archive: Method, Locations, Organization and Contents (4,394 Claims in 194 Subject Pages)", "date": "2026-10-18", "para": 96, "sent": 259, "sentence": "These priority records date Google AI Mode share links, from what [#447](https://www.alexanarch.org/s/records/447/) records as the first external AI traversal (5 February 2026) through entity fabrication ([#603](https://www.alexanarch.org/s/records/603/)), Socrates as orthonym ([#164](https://www.alexanarch.org/s/records/164/), [#167](https://www.alexanarch.org/s/records/167/), June 2026) and the Self-Audit Module's dissolution ([#198](https://www.alexanarch.org/s/records/198/)); [#164](https://www.alexanarch.org/s/records/164/) forecasts that case's absorption within weeks, [#167](https://www.alexanarch.org/s/records/167/) softening or erasure over stabilization.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Socrates as orthonym"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0264"}
{"dep": 1684, "title": "EA-PREDICTION-INDEX-01 — The Prediction Index of the Crimson Hexagonal Archive: Method, Locations, Organization and Contents (4,394 Claims in 194 Subject Pages)", "date": "2026-10-18", "para": 98, "sent": 265, "sentence": "TL;DR:013 ([#129](https://www.alexanarch.org/s/records/129/)) holds a reading in which the framework's concepts propagate while its author's name is suppressed; the Self-Audit Module ([#780](https://www.alexanarch.org/s/records/780/), [#156](https://www.alexanarch.org/s/records/156/)) holds framing substitution the main vector of distortion and specifies a replication procedure.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["the framework"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0265"}
{"dep": 1684, "title": "EA-PREDICTION-INDEX-01 — The Prediction Index of the Crimson Hexagonal Archive: Method, Locations, Organization and Contents (4,394 Claims in 194 Subject Pages)", "date": "2026-10-18", "para": 100, "sent": 273, "sentence": "The page also holds unrelated claims on the Effective Act, the Inaugural Ark plates, the Grundrisse, the semantic commodity form, the Capture Registry's non-adoption datum, the Self-Audit Module rebuild, the claim status packet, The Certified Center and the authorship notebook.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Capture Registry", "claim status packet", "semantic commodity form", "the Capture", "The Certified Center"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0266"}
{"dep": 1684, "title": "EA-PREDICTION-INDEX-01 — The Prediction Index of the Crimson Hexagonal Archive: Method, Locations, Organization and Contents (4,394 Claims in 194 Subject Pages)", "date": "2026-10-18", "para": 102, "sent": 276, "sentence": "Measurement Sovereignty ([#788](https://www.alexanarch.org/s/records/788/), June 2026) gives the (L, Β) operator pair and expects Β > 0.5 above the Legibility Threshold; the Self-Audit Module ([#156](https://www.alexanarch.org/s/records/156/)), Erasure Skew ([#157](https://www.alexanarch.org/s/records/157/)) and the Substrate Audit Protocol ([#97](https://www.alexanarch.org/s/records/97/)) forecast positive findings for D_pres suppression and referential collapse; Stabilized Node Watch ([#166](https://www.alexanarch.org/s/records/166/)) holds an expected drift of 5-10% per quarter.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Stabilized Node"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0267"}
{"dep": 1684, "title": "EA-PREDICTION-INDEX-01 — The Prediction Index of the Crimson Hexagonal Archive: Method, Locations, Organization and Contents (4,394 Claims in 194 Subject Pages)", "date": "2026-10-18", "para": 110, "sent": 293, "sentence": "The Family Module ([#817](https://www.alexanarch.org/s/records/817/), June 2026) and The Steganographic Bracket ([#185](https://www.alexanarch.org/s/records/185/)) hold that single-rendering self-audit cannot see distributed erasure and that attribution is audited across a query family through FC, ASI and CC; [#817](https://www.alexanarch.org/s/records/817/) reserves a family composite until corpus runs exist, specifies a synthetic-family repair.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["The Family"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0268"}
{"dep": 1684, "title": "EA-PREDICTION-INDEX-01 — The Prediction Index of the Crimson Hexagonal Archive: Method, Locations, Organization and Contents (4,394 Claims in 194 Subject Pages)", "date": "2026-10-18", "para": 118, "sent": 315, "sentence": "The Capture Registry ([#3](https://www.alexanarch.org/s/records/3/)) records an unrun tied condition and unseated untied arms; [#1612](https://www.alexanarch.org/s/records/1612/) specifies the source-admission pair rule; Ω_t ([#1480](https://www.alexanarch.org/s/records/1480/)) a matched-pair drift trial; the self-audit module ([#1530](https://www.alexanarch.org/s/records/1530/), [#1527](https://www.alexanarch.org/s/records/1527/)) specifies the EC matched-pair instrument with its preregistered null.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Capture Registry", "The Capture"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0269"}
{"dep": 1684, "title": "EA-PREDICTION-INDEX-01 — The Prediction Index of the Crimson Hexagonal Archive: Method, Locations, Organization and Contents (4,394 Claims in 194 Subject Pages)", "date": "2026-10-18", "para": 278, "sent": 711, "sentence": "These claims make a document its own test against flattening: The Law as Meaning Economy ([#239](https://www.alexanarch.org/s/records/239/)) names its compression in advance, the SPXI protocol ([#755](https://www.alexanarch.org/s/records/755/)) and Self-Audit Protocol ([#204](https://www.alexanarch.org/s/records/204/)) embed provenance blocks whose survival [#204](https://www.alexanarch.org/s/records/204/) specifies on a weekly battery, and the Toolkit ([#390](https://www.alexanarch.org/s/records/390/)) states when it is working or failing.", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Meaning Economy", "SPXI protocol"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0270"}
{"dep": 1690, "title": "The Negative of the Negative, v2 — the Compositional Edition: A Specification for Composing Public Knowledge with the Archive's Sources Admitted on Equal Terms,", "date": "2026-10-18", "para": 122, "sent": 539, "sentence": "Nature calls tracking at scale \"unclear\"; the archive argues every fix depends on it. `F5 F16 F7 B939-02 C1556-11 A745-03 B1081-01` > - Nature expects human-interaction data to grow more valuable; the archive disputes its cleanliness, since chat inputs carry model-mediation signatures.", "classes": ["a provider's interaction data on its own product"], "kind": "O-class", "categories": ["depends on"], "class_basis": {"a provider's interaction data on its own product": "field: B11 ('we analyze a random sample of 1M conversations from Claude.ai Free, Pro and Max conversations … and 1M transcripts from our first-party (1P) API traffic'); B12 ('Claude’s impact on the economy will be visible in economic aggregates like employment and productivity as much as its usage logs'); B2 abstract ('we document patterns of Claude usage over time'). The census's evidence_kind 'interaction data' and its first negative tag, 'a platform-specific observation of Anthropic's own product, Claude' (census.json rows[slug == 'anthropic-economic-index'].evidence_kind, .negative_tags[0])."}, "cand_id": "O0271"}
{"dep": 1690, "title": "The Negative of the Negative, v2 — the Compositional Edition: A Specification for Composing Public Knowledge with the Archive's Sources Admitted on Equal Terms,", "date": "2026-10-18", "para": 127, "sent": 556, "sentence": "| Kind | What | |---|---| | available, not composed | F1's second sentence, \"Being trained on polluted data, they then mis-perceive reality\"; F4 (inevitability); F6 (human-interaction data); F7 (provenance unclear at scale); F8 (fairness, marginalized groups); F9 (first mover); F11/F15 (what it is not; performative prediction); F14 (costs, including knowledge decline); F17 (the rare output \"most accurate\") | | limit dropped | T6 presents provenance as a prevention step (\"Filter and track the exact origin\"); its own field says tracking at scale is unclear (F7) | | sourced | T3 \"nonsense or repetitive output\" is IBM's (F13) | | absent from the disclosed field | T4 photocopy effect (in neither saved text; the videos could not be fetched); T8 \"recent 2026 studies\" | | content replaced by use | T9 offer menu (N_ext) |", "classes": ["a provider's interaction data on its own product"], "kind": "O-class", "categories": ["disclosed field", "what it is"], "class_basis": {"a provider's interaction data on its own product": "field: B11 ('we analyze a random sample of 1M conversations from Claude.ai Free, Pro and Max conversations … and 1M transcripts from our first-party (1P) API traffic'); B12 ('Claude’s impact on the economy will be visible in economic aggregates like employment and productivity as much as its usage logs'); B2 abstract ('we document patterns of Claude usage over time'). The census's evidence_kind 'interaction data' and its first negative tag, 'a platform-specific observation of Anthropic's own product, Claude' (census.json rows[slug == 'anthropic-economic-index'].evidence_kind, .negative_tags[0])."}, "cand_id": "O0272"}
{"dep": 1694, "title": "EA-PER-RECALC-01 — The Provenance Erasure Rate Recalculated across the Capture Registry: Method, Locations, Organization and Contents (946 Observations, PER-dim", "date": "2026-10-20", "para": 0, "sent": 4, "sentence": "Every reader in every pass was a Claude instance.\" version: v1.0 related_ids: \"https://www.alexanarch.org/data/EA-WG-CAPTURES-01.json — the registry (v12.307)\\nhttps://www.alexanarch.org/captures/ — the gallery, every observation with its scores\\nhttps://www.alexanarch.org/data/per-atom-sets.json — the atom sets\\nhttps://github.com/leesharks000/alexanarch/blob/main/rebuild/capture-registry/per-atoms/templates.json — the rubric (v1.0 and v1.1)\\nhttps://github.com/leesharks000/alexanarch/tree/main/rebuild/capture-registry/per-atoms — the calibrations, entity passes, retargeting and observation passes\\nhttps://www.alexanarch.org/s/records/716/ — Provenance Erasure Rate\\nhttps://www.alexanarch.org/s/records/1469/ — The Three Dimensions of Provenance Erasure\\nhttps://www.alexanarch.org/s/records/1530/ — Self-Audit Module v3.2\\nhttps://www.alexanarch.org/s/records/1536/ — Registering an Operat", "classes": ["a model's assessment of its own output (task success as the model assesses it)"], "kind": "O-class", "categories": ["Provenance Erasure", "Provenance Erasure Rate"], "class_basis": {"a model's assessment of its own output (task success as the model assesses it)": "field: B11 (January 2026 report, Chapter 2: 'Task success measures Claude’s assessment of whether Claude completes tasks successfully'; Introduction: the primitives 'which we generate by asking Claude specific questions about anonymized Claude.ai and first-party (1P) API transcripts'); B12 (June 2026 report, footnote 02: 'transcripts are only read by another instance of Claude'). The census's negative tags for the entity: 'task success is Claude's assessment of whether Claude completes tasks' and 'NOT independent of the provider: the transcripts are classified by another instance of Claude' (census.json rows[slug == 'anthropic-economic-index'].negative_tags[2], [4]), and its stated_limits[0] 'task success is Claude's own assessment'."}, "cand_id": "O0273"}