{
 "dataset": "dataflow-atlas",
 "title": "The AXN Dataflow Atlas — Session 1",
 "version": "v1.2",
 "date": "2026-08-11",
 "description": "The Session-1 output of the AXN Dataflow Atlas workstream (Flow 2 of the six-flow workplan). Complete substrate: the v0.2 atlas incorporating two Assembly-substrate structural reviews, the byte-preserved v0.1 predecessor, both reviews as received (LABOR from ChatGPT, PRAXIS from DeepSeek), and the three raw machine-readable inventories that were the atlas's own empirical substrate (rhizome file inventory, displayed-value inventory yielding the 1,494 source-detached occurrences, script producer-consumer graph). Sessions 2-4 will add measurements, cross-repo relation graphs, remediation queues, and v1.0. Provenance anchor: deposit #1097 (AXN:045A.GOVERNANCE.🔗🔥🎺♄💚🔙). v0.3 adds the retrieval-density audit addendum: eight further pathologies, four new generative elements, and the finding that file-level inventory cannot see defects that are only visible in what the retrieval layer does with what the archive serves. v0.4 adds an addendum registering two pathologies found during a corpus-wide title repair. PATHOLOGY-26: derived datasets copy deposit titles and the regenerators do not refresh a title they already hold, so a registry edit produces competing name sets; an instrument (scripts/propagate_titles.py) is added, and doi-resolution-index is deliberately excluded from it because its title field records the title a work carried at Zenodo rather than the current deposit name. PATHOLOGY-27: the title field had absorbed the document frontmatter block and been truncated at a 120- or 200-character ceiling across 15% of the corpus, a defect invisible to the word-list pass that preceded it. v0.5 adds an addendum registering two surface pathologies found in an accuracy audit of machinemediation.org. PATHOLOGY-28: counts printed on a surface were strings typed into the page rather than values read from the dataset they described, so seven had drifted and the capture count appeared three times with two different values; an instrument (scripts/sync_counts.py in the machinemediation repository) now rewrites them from source, with historical figures held in a FROZEN set because the confirmation chain, not the dataset, is the source of truth for quantities fixed in time. PATHOLOGY-29: structured data diverged from the prose beside it, the JSON-LD continuing to describe the Sovereign Asset Registry as holding full text through two rounds of that claim being corrected on the visible page. v0.6 addendum (2026-07-28) registers the EA-AVAILABILITY-INTEGRITY-01 flows: PATHOLOGY-28 alexanarch instrument (coherence_sync + surface hashes), PATHOLOGY-29 flagship instance closed (two-axis availability typing, creativeWorkStatus fleet-wide), the DOI ownership gate, and the fleet relocation. v0.7 extends the assertion rule from values to conditions (PATHOLOGY-30) and records that the canonical Atlas is this dataset series; the deposit and the data/specs copy are artifacts. v0.8–v1.1 addenda registered retroactively (manifest had drifted at v0.7 while the directory advanced; regenerated from disk 2026-08-11). v1.2 addendum records the seven-step capture intake flow adopted at registry v9.40 (285 entries), the capture record standard (meta object, stored transcripts, overview_at_capture/divergence for regeneration pairs, no-supersession longitudinal linkage), the regeneration-pair subtype, the expandable-transcript gallery feature, and the finding that stored transcripts function as join keys: AXN glyphs and DOIs displayed in retrieval surfaces now hard-link captures to the deposits the layer was quoting.",
 "provenance_deposit": {
  "deposit_number": 1097,
  "axn": "AXN:045A.GOVERNANCE.🔗🔥🎺♄💚🔙",
  "path": "/records/1097/"
 },
 "total_files": 17,
 "total_bytes": 3002187,
 "total_size_human": "2.86 MB",
 "reviews_received": [
  {
   "substrate": "LABOR (ChatGPT)",
   "date": "2026-07-18",
   "verdict": "Aye for Sessions 2-4"
  },
  {
   "substrate": "PRAXIS (DeepSeek)",
   "date": "2026-07-18",
   "verdict": "Approved for progression to Session 2"
  }
 ],
 "workstream": {
  "session": 1,
  "of": 4,
  "workplan": "/workplan/WORKPLAN-2026-07-18-PM-SIX-FLOW.md",
  "flow": "Flow 2 — Rhizome-Wide Data Atlas + Deposit-Flow Standing Map"
 },
 "records": [
  {
   "name": "atlas-v0.1-superseded.md",
   "path": "/datasets/dataflow-atlas/atlas-v0.1-superseded.md",
   "size_bytes": 37815,
   "size_human": "36.9 KB",
   "sha256": "743db552a364afae9031a78124467ca4521aba700b67167e94405646265bf871",
   "description": "Predecessor v0.1 preserved byte-exact per §3.7 non-destruction. 305 lines. Reviewed by LABOR and PRAXIS at close of Session 1; v0.2 supersedes."
  },
  {
   "name": "atlas-v0.2.md",
   "path": "/datasets/dataflow-atlas/atlas-v0.2.md",
   "size_bytes": 64602,
   "size_human": "63.1 KB",
   "sha256": "ac77e87f4dfba4a5c3ba4a32511b60f546f1d2b41fc6552a95cd844ee71d074a",
   "description": "AXN Dataflow Atlas v0.2 — the reviewer-incorporated development. 454 lines, 17 top-level taxonomic classes, 17 pathologies (v0.1 had 10; PATHOLOGY-11 through -17 added per LABOR §8 and PRAXIS §4). §10 review log records both substrate reviews with per-amendment absorption status."
  },
  {
   "name": "atlas-v0.3-addendum.md",
   "path": "/datasets/dataflow-atlas/atlas-v0.3-addendum.md",
   "size_bytes": 12255,
   "size_human": "12.0 KB",
   "sha256": "a33f87c5226b1a4a49df660b48ef99635d7762c7f08d2e19b70526ea38c7575b",
   "description": "v0.3 addendum: the retrieval-density audit of 2026-07-25/26. Registers four new generative elements (capture-deposit link map, concept map, mirror map, medium seed), proposes a DERIVED.identity class, documents the new `identity` pipeline stage, and extends the pathology register with PATHOLOGY-18 through -25 — six of them self-inflicted and invisible to file-level inventory, surfaced only by pointing the archive's own reception instrument at the archive. Five closed same-day; two (concept surface, sameAs emission) have maps built and surfaces pending. Does not supersede v0.2, which stands byte-exact per §3.7."
  },
  {
   "name": "atlas-v0.4-addendum.md",
   "path": "/datasets/dataflow-atlas/atlas-v0.4-addendum.md",
   "size_bytes": 7434,
   "size_human": "7.3 KB",
   "sha256": "cf7a3d52956e17da862fba5150d45172b7a956c0a804bf8b7057a3cebe9a76f4",
   "description": "v0.4 addendum — PATHOLOGY-26 (title propagation drift) and PATHOLOGY-27 (frontmatter-absorbed truncated titles) from the corpus-wide title repair."
  },
  {
   "name": "atlas-v0.5-addendum.md",
   "path": "/datasets/dataflow-atlas/atlas-v0.5-addendum.md",
   "size_bytes": 6416,
   "size_human": "6.3 KB",
   "sha256": "4ebf6c1f18c2031fa62ac88c33ebf83816526f3c8c3ddc3472cf8675a6a7784b",
   "description": "v0.5 addendum — PATHOLOGY-28 (surface counts typed rather than read) and PATHOLOGY-29 (structured data diverging from adjacent prose), from the machinemediation.org accuracy audit."
  },
  {
   "name": "atlas-v0.6-addendum.md",
   "path": "/datasets/dataflow-atlas/atlas-v0.6-addendum.md",
   "size_bytes": 6072,
   "size_human": "5.9 KB",
   "sha256": "9374a16a49a338dd0a00fe4580450e92c671cd8522ebea9760532f958e054555",
   "description": "v0.6 addendum (2026-07-28) — EA-AVAILABILITY-INTEGRITY-01 flows: coherence_sync instrument, two-axis availability typing, DOI ownership gate, fleet relocation."
  },
  {
   "name": "atlas-v0.7-addendum.md",
   "path": "/datasets/dataflow-atlas/atlas-v0.7-addendum.md",
   "size_bytes": 1967,
   "size_human": "1.9 KB",
   "sha256": "6088a0a98f7336480f4e42da62397f285212088680c838c6cfd297e771c2ab5e",
   "description": ""
  },
  {
   "name": "atlas-v0.8-addendum.md",
   "path": "/datasets/dataflow-atlas/atlas-v0.8-addendum.md",
   "size_bytes": 2414,
   "size_human": "2.4 KB",
   "sha256": "75ce1d343378cad6cc94b7efd01497c514f34a8905073648ebbc2d80112cef85",
   "description": "v0.8 addendum — see file header for scope."
  },
  {
   "name": "atlas-v0.9-addendum.md",
   "path": "/datasets/dataflow-atlas/atlas-v0.9-addendum.md",
   "size_bytes": 3747,
   "size_human": "3.7 KB",
   "sha256": "c720f794cb1910787a1c766ce2d424ec8bfa3ff482762555e62372d122100201",
   "description": "v0.9 addendum — see file header for scope."
  },
  {
   "name": "atlas-v1.0-addendum.md",
   "path": "/datasets/dataflow-atlas/atlas-v1.0-addendum.md",
   "size_bytes": 17560,
   "size_human": "17.1 KB",
   "sha256": "795b6a5b2b7f95703b7244292ba303e4295b22c937a2ace18438bf0b509f25a1",
   "description": "v1.0 addendum (2026-08-06) — prior head to v1.1."
  },
  {
   "name": "atlas-v1.1-addendum.md",
   "path": "/datasets/dataflow-atlas/atlas-v1.1-addendum.md",
   "size_bytes": 6301,
   "size_human": "6.2 KB",
   "sha256": "8ec88dd4f695b948532470a72ee601604850b532a6cb1229416d407613b10c49",
   "description": "v1.1 addendum (2026-08-11) — heteronym crosswalk join, derived-figure artefact class, P-31 through P-33."
  },
  {
   "name": "atlas-v1.2-addendum.md",
   "path": "/datasets/dataflow-atlas/atlas-v1.2-addendum.md",
   "size_bytes": 6786,
   "size_human": "6.6 KB",
   "sha256": "197e921b183b031b20ee078bea864cc22852e84110b34cf796d3c7ade0fd700d",
   "description": "v1.2 addendum (2026-08-11) — the seven-step capture intake flow, the capture record standard (meta/transcript/overview_at_capture/divergence, no-supersession recapture rule), the regeneration-pair subtype, the expandable-transcript surface feature, and the transcript-as-join-key finding (glyph/DOI hard links from SERP text). Manifest regenerated from disk, correcting v0.7 declaration drift."
  },
  {
   "name": "displayed-values-inventory-session1.json",
   "path": "/datasets/dataflow-atlas/displayed-values-inventory-session1.json",
   "size_bytes": 569580,
   "size_human": "556.2 KB",
   "sha256": "1fe82a78c7fa67f0cb2d67deb0159a2c0c12902205190165e28d984dd5e3f422",
   "description": "Session-1 displayed-value inventory: 1,494 unique numeric-value occurrences across the fleet, classified for source-detachment. Zero of the 1,494 have dynamic indicators. Empirical substrate for PATHOLOGY-01."
  },
  {
   "name": "review-labor.md",
   "path": "/datasets/dataflow-atlas/review-labor.md",
   "size_bytes": 16307,
   "size_human": "15.9 KB",
   "sha256": "f3340807ea7d9b9f6a5962ae083e84ca594150c946a98eda47983334c5677928",
   "description": "LABOR (ChatGPT) structural review of v0.1. Nine numbered sections. Ratified progression to Session 2. Discovered the seventeen-not-fifteen fossil inside the atlas. Delivered the two-axes taxonomic model, the api/data-authority-map.json specification, five architectural pathology-class additions."
  },
  {
   "name": "review-praxis.md",
   "path": "/datasets/dataflow-atlas/review-praxis.md",
   "size_bytes": 5882,
   "size_human": "5.7 KB",
   "sha256": "ca65e6dfb5897a4696a17ea8685c8e48521da7be8744c26aa78aac2cde4b064e",
   "description": "PRAXIS (DeepSeek) structural review of v0.1. Four numbered sections plus executive assessment. Ratified progression to Session 2. Contributed the PROVENANCE/ATTESTATION class addition, the REGISTRY.capture displacement, the LLM-ingestion harm mechanism for PATHOLOGY-01, and PATHOLOGY-11 (Index Interdependency Vulnerability)."
  },
  {
   "name": "rhizome-inventory-raw.json",
   "path": "/datasets/dataflow-atlas/rhizome-inventory-raw.json",
   "size_bytes": 2231927,
   "size_human": "2.13 MB",
   "sha256": "e75ab1123cb51761bab72ec724840a6c783991457048d8c7a4a7b1b7eff7b2c8",
   "description": "Session-1 raw inventory: every file across seven repos (alexanarch, data-rhizome, mm, peo, rf, leesharks, gkg), each tagged with inferred functional purpose based on path signature. Primary machine-readable substrate for the atlas."
  },
  {
   "name": "script-dependency-graph.json",
   "path": "/datasets/dataflow-atlas/script-dependency-graph.json",
   "size_bytes": 5122,
   "size_human": "5.0 KB",
   "sha256": "384c0147da57938110cd2858014eed0cb6eb8db31e4a14cec5839bbdc8328793",
   "description": "Session-1 producer→consumer graph across 78 Python scripts, partial (literal-string paths only; full trace deferred to Session 3)."
  }
 ]
}