Bound Variovorax causal graph and correct strain-aware census (batch 43)

history/records/GLBRC_Populus_Variovorax_SynCom28/2026-10-06T113342Z-codex-0ba641.yaml

@@ -0,0 +1,27 @@
+history_version: 1
+target:
+  kind: record
+  path: kb/communities/GLBRC_Populus_Variovorax_SynCom28.yaml
+  slug: GLBRC_Populus_Variovorax_SynCom28
+session:
+  id: 2026-10-06T113342Z-codex-0ba641
+  timestamp: '2026-10-06T11:33:42Z'
+  actors:
+  - type: ai_agent
+    name: codex
+    model: unspecified
+    agent_tool: codex
+links:
+  issues:
+  - https://github.com/CultureBotAI/CommunityMech/issues/1471
+events:
+- type: EDIT
+  outcome: changed
+  sections:
+  - interactions
+  - discussions
+  summary: Retain relative compartment enrichment, not reduced competition or absolute colonization
+    efficiency.; Retain author-proposed resource competition as HYPOTHESIZED without a selective
+    compartment-causation claim.; Restrict the existing fucose node to canonical CF313 locus-associated
+    root recovery, not interstrain facilitation or pathway sufficiency.
+  details: Source-based self-adversarial repairs to existing graph; no new structure. reports/causal_graph_review/decisions/20261006-variovorax-batch43.yaml

reports/causal_graph_review/README.md

@@ -5,6 +5,17 @@ source-based self-adversarial review. These records are not independent approval
 or proof that every graph is complete. The current inventory and summary distinguish
 `reviewed`, `pending` and `needs_research`.
 
+## Connectivity Census Correction (2026-10-06)
+
+The historical test survey unioned names, labels and CURIEs for pairwise endpoints.
+This could credit every strain sharing a genus ID even when an endpoint named only
+one strain. Batch43 reuses the already strain-aware readiness resolver; production
+auditor behavior is unchanged. Historical raw survey outputs remain unchanged, but
+their `credited_solely_by_the_rule` counts must not be interpreted as exact
+strain-resolved counts. [Batch43 validation](validation-20261006-batch43.json)
+records both algorithms on both the pre-edit and post-edit corpus, separating this
+measurement correction from the CF313 scope change. See [issue 1472](https://github.com/CultureBotAI/CommunityMech/issues/1472).
+
 ## Reference-Validator Correction (2026-10-06)
 
 Some historical batch reports incorrectly interpret `Total checks: 0` as zero

reports/causal_graph_review/adversarial-review-20261006-batch43.yaml

@@ -0,0 +1,150 @@
+reviewer: codex
+review_mode: source_based_self_adversarial_review
+independent_approval: false
+base_commit: 22b5b12c2386a949d1ec4ddb1f79ee288085bd1c
+scope:
+- kb/communities/GLBRC_Populus_Variovorax_SynCom28.yaml
+decision_ledger: reports/causal_graph_review/decisions/20261006-variovorax-batch43.yaml
+validation_report: reports/causal_graph_review/validation-20261006-batch43.json
+findings:
+- issue: 1471
+  severity: P2
+  status: addressed_in_this_change
+  finding: Relative community enrichment and a separate CF313 mutant result were overstated as competitive
+    partitioning and interstrain facilitation.
+  disposition: Three existing nodes bounded/renamed, both unsupported arrows removed, exact canonical
+    participants, focused gap and history. No new graph structure. Close only after protected main merge.
+- issue: 1472
+  severity: P2
+  status: addressed_in_this_change
+  finding: Legacy test census credited every shared-genus strain to a precisely named single-strain endpoint.
+  disposition: Reuse existing exact-name-aware ranker resolver; production auditor/ranker unchanged. Compare
+    both algorithms on both corpus versions, retain synthetic shared-CURIE canaries and add live CF313
+    test. Historical counts retained but not certified as exact strain counts. Close only after main merge.
+adversarial_checks:
+- Whole 1,210-line original record and all three nodes/two arrows read. All 28 canonical members retain
+  their strain preferred names with the shared genus CURIE; no invented strain/species IDs or genus self-edge.
+- Complete cached primary main body, Methods and embedded captions read despite stale abstract_only front
+  matter. Primary public PMC request presented reCAPTCHA, with no bypass. No figure images, supplementary
+  files, raw data, code or complementation rescue independently inspected.
+- Compartment metagenomics measures relative DNA enrichment and normalized abundance, not absolute viable
+  colonization or experimentally reduced competition. A draft absolute-read-abundance phrase was caught
+  during self-review, corrected before commit and followed by fresh complete verification. Initial run
+  and its hashes remain archived.
+- Resource competition remains HYPOTHESIZED and evidence PARTIAL. The source proposes alternative metabolic/regulatory
+  strategies but does not selectively measure resource-mediated partner harm. Shared genus is not proof
+  of a single species; no intraspecific category retained.
+- CF313 PMI12_01353 deletion abolishes sole-fucose growth and reduces recovered root CFUs in P. deltoides,
+  reported P=0.0022. This existing node is source-only CF313; not partner facilitation, priority effects,
+  whole-panel compartment causation or proven entry/flux mechanism. Other intact-pathway strains fail
+  growth; pathway possession alone is not sufficient. Host fucose not quantified.
+- Both downstream arrows removed because their targets represent relative community compartment profiles
+  rather than the separate mutant root-recovery endpoint. No new causal nodes/arrows, no Edison job or
+  provider spend. Canonical L-fucose grounding reused from this record.
+- The historic all-community-level auditor fixture is now explicitly synthetic, preserving exact-name
+  versus ambiguous-CURIE behavior tests without freezing scientific data. The live record has two full-panel
+  observations and one intrinsic source. New test verifies one of 28 precise matches and no fan-out for
+  ambiguous ID-only pairwise endpoints.
+- 'Census legacy/current and corrected/baseline comparisons distinguish algorithm correction from biology.
+  Mixed count changes 124 to 125 and CL-only 263 to 262; only that bound changes. Historical raw numeric
+  outputs are not rewritten or represented as exact strain-aware counts. README and issue document correction;
+  broader #524 remains open.'
+- 459 other records, 920 source caches, taxonomy and all non-graph raw blocks preserved. Prior history
+  artifacts byte-identical; one append-only event/artifact. Both initially selected UFMP/GOM records remain
+  pending, not silently counted as reviewed.
+- Schema, strict semantics, term correspondence, reference, snippet, history and rank checks passed on
+  final bytes. No new audit findings; all 33 baseline warnings preserved, zero structural errors. Repeat
+  renders deterministic and all published node/edge descriptions checked. No browser screenshot claimed.
+- 'Five new focused tests plus the combined historical batch/shared suite pass. The initial run passed
+  445 tests but failed Ruff on import ordering and dict-constructor style; fixes and its actual failure
+  output are retained, followed by a fresh complete run. Positive/negative reference-validator canaries
+  remain included. Total checks:0 is an issue count, not a no-op; SupportingReference #476 remains incomplete.
+  Text matching is not source-grounding validation.'
+- All-state 788 issue title/body snapshot and ignored/hidden local search used. Relevant biological and
+  census matches read; initial mistaken data/history scan corrected to history/ before concluding dedup.
+  Not every issue comment or PR review inspected.
+- Self-adversarial review is not independent approval. Exact-head COMMENT and actual post-commit docs-current
+  are publication steps; protected main checks and dependent-parent merges remain mandatory. Neither complete
+  corpus review nor completed merge is asserted here.
+verification:
+  combined_regressions: 445 passed in 176.93s (0:02:56)
+  new_regression_tests: 5
+  schema_strict_terms_history: passed with actual outputs and digests
+  references: 'CLI Total checks counts reported issues, not executed checks. This record exits 0 with
+    no reported issues; 44 snippet-audit matches are textual matching, not proof of scientific support
+    or independently verified provenance. SupportingReference coverage remains incomplete under #476;
+    earlier counter interpretation is corrected by batch42.'
+  snippet_counts_by_record:
+    kb/communities/GLBRC_Populus_Variovorax_SynCom28.yaml:
+      MATCH: 44
+      RENDERING: 0
+      ASSEMBLED: 0
+      WEAK: 0
+      MISMATCH: 0
+      NOCONTENT: 0
+  network: Zero errors; exactly 33 existing warnings, none added.
+  black_ruff_diff: passed
+  tested_file_sha256:
+    reports/causal_graph_review/decisions/20261006-variovorax-batch43.yaml: 8740238b2df4bbb42e8cdceb7e8ca47dd1734b96a96fc161b766f5ccb7973f41
+    tests/test_causal_review_batch43.py: 3444fc98c3cf63e9dcaf90de30f761252799087b74821b3ac155b995ec138afe
+    tests/test_participating_taxa.py: c78bbb89bc2afc5c6a9e646bb9c5103e1cbb22d0f19b09f9fa7001e7ed2ba8a5
+    tests/test_community_level_connectivity_credit.py: a9ba9fbacb9edaffbd1590ecc2f8fbc1b4ab36a3caf358f90c943c8a64fd76fd
+    kb/communities/GLBRC_Populus_Variovorax_SynCom28.yaml: c251f6d716591cd391dbb09ad961f6ccb7b123988b9c677d925356b8463d2c28
+    reports/causal_graph_review/README.md: 494d79788f9b4c63def8d22ef421081ea2facff92a709d5ca577b4ede46958f3
+  post_commit_docs_current: Separate publication gate, not asserted before commit.
+participant_census:
+  measurements:
+    corrected_baseline:
+      community_level_only: 263
+      credited_solely_by_the_rule: 1222
+      mixed: 124
+      records: 460
+      taxa: 1494
+      with_community_level: 387
+    corrected_current:
+      community_level_only: 262
+      credited_solely_by_the_rule: 1221
+      mixed: 125
+      records: 460
+      taxa: 1494
+      with_community_level: 387
+    legacy_baseline:
+      community_level_only: 263
+      credited_solely_by_the_rule: 1219
+      mixed: 124
+      records: 460
+      taxa: 1494
+      with_community_level: 387
+    legacy_current:
+      community_level_only: 262
+      credited_solely_by_the_rule: 1191
+      mixed: 125
+      records: 460
+      taxa: 1494
+      with_community_level: 387
+  correction_issue: 1472
+  scientific_issue: 1471
+  interpretation: Production auditor/ranker unchanged. Legacy survey falsely credits all 28 shared-genus
+    members to CF313; corrected algorithm credits exactly CF313. Historical survey outputs are retained
+    but withdrawn as exact strain-resolved counts. One new participant record, 189 to 190. Mixed-scope
+    ceiling increases only 124 to 125; other bounds unchanged.
+corpus:
+  records: 460
+  nodes: 1468
+  edges: 684
+  review_status_counts:
+    needs_research: 12
+    pending: 284
+    reviewed: 164
+limitations:
+- Complete cached primary main body, Methods and embedded captions read despite stale abstract_only front
+  matter; no figure images, separate supplements, raw data or analysis code independently inspected. Public
+  PMC browser path presented reCAPTCHA; no bypass.
+- Three existing nodes retained/renamed, two unsupported arrows removed. No new causal structure or Edison
+  spend.
+- Canonical taxonomy, non-graph raw blocks, historical records and all original caches preserved, not
+  certified. Both initially selected UFMP/GOM candidates remain pending.
+- 'The reference CLI Total checks counter counts reported issues, not executed checks; SupportingReference
+  coverage remains incomplete under #476. Text matching is not scientific support or independent cache
+  provenance.'
+- Source-based self-adversarial review is not independent approval or corpus completion.

reports/causal_graph_review/decisions/20261006-variovorax-batch43.yaml

@@ -0,0 +1,101 @@
+reviewer: codex
+review_mode: source_based_self_adversarial_review
+independent_approval: false
+base_commit: 22b5b12c2386a949d1ec4ddb1f79ee288085bd1c
+baseline_commit: 22b5b12c2386a949d1ec4ddb1f79ee288085bd1c
+base_tree: 71133d5dcc9a140865ceb58ccfe7b26125738c40
+issues:
+- 1471
+open_nongraph_followups:
+- 524
+issue_deduplication:
+  all_state_issue_titles_and_bodies: 788
+  ignored_hidden_local_search: true
+  relevant_bodies_read:
+  - 1251
+  - 762
+  - 615
+  - 524
+  - 485
+  - 312
+  - 18
+  - 12
+  limitations: Not all issue comments or PR reviews inspected; initial nonexistent data/history scan corrected to
+    history/ before dedup conclusion.
+reference_cache_original_hashes:
+  references_cache/PMID_41543249.md: 8efa077f04918a0c73cc55cb4dcb14326c01c2a62b6a41ef727f340b19e40ec6
+limitations:
+- Complete cached primary main body, Methods and embedded captions read despite stale abstract_only front matter;
+  no figure images, separate supplements, raw data or analysis code independently inspected. Public PMC browser
+  path presented reCAPTCHA; no bypass.
+- Three existing nodes retained/renamed, two unsupported arrows removed. No new causal structure or Edison spend.
+- Canonical taxonomy, non-graph raw blocks, historical records and all original caches preserved, not certified.
+  Both initially selected UFMP/GOM candidates remain pending.
+- 'The reference CLI Total checks counter counts reported issues, not executed checks; SupportingReference coverage
+  remains incomplete under #476. Text matching is not scientific support or independent cache provenance.'
+- Source-based self-adversarial review is not independent approval or corpus completion.
+records:
+- path: kb/communities/GLBRC_Populus_Variovorax_SynCom28.yaml
+  id: CommunityMech:000028
+  original_sha256: e6b9446f0f493eb1c8b2666dae051146e4375795c7d5bf51f485d09cb1e8170d
+  record_sha256: c251f6d716591cd391dbb09ad961f6ccb7b123988b9c677d925356b8463d2c28
+  status: reviewed
+  outcome: changed
+  review_file: reports/causal_graph_review/decisions/20261006-variovorax-batch43.yaml
+  scope: Whole record and all three original nodes/two arrows reviewed; source and scientific limits explicit.
+  summary: Retain relative compartment enrichment, not reduced competition or absolute colonization efficiency.;
+    Retain author-proposed resource competition as HYPOTHESIZED without a selective compartment-causation claim.;
+    Restrict the existing fucose node to canonical CF313 locus-associated root recovery, not interstrain facilitation
+    or pathway sufficiency.
+  renamed_nodes:
+    Rhizosphere-Endosphere Niche Partitioning: Variovorax Relative Compartment Enrichment
+    Intraspecific Metabolic Competition: Proposed Variovorax Resource Competition
+    L-Fucose Pathway Colonization Enhancement: CF313 Fucose-Pathway-Associated Root Recovery
+  node_decisions:
+  - node: Variovorax Relative Compartment Enrichment
+    rationale: Retain relative compartment enrichment, not reduced competition or absolute colonization efficiency.
+  - node: Proposed Variovorax Resource Competition
+    rationale: Retain author-proposed resource competition as HYPOTHESIZED without a selective compartment-causation
+      claim.
+  - node: CF313 Fucose-Pathway-Associated Root Recovery
+    rationale: Restrict the existing fucose node to canonical CF313 locus-associated root recovery, not interstrain
+      facilitation or pathway sufficiency.
+  removed_node_decisions: []
+  edges_before:
+  - source: Intraspecific Metabolic Competition
+    target: Rhizosphere-Endosphere Niche Partitioning
+    description: Competition is one of the ecological strategies that lets Variovorax strains succeed during Populus
+      endosphere colonization and produces distinct, host-filtered rhizosphere and endosphere assemblages.
+  - source: L-Fucose Pathway Colonization Enhancement
+    target: Rhizosphere-Endosphere Niche Partitioning
+    description: The conserved L-fucose utilization pathway enhances root colonization and gives pathway-bearing
+      strains an experimentally validated route through the endosphere selection bottleneck.
+  edges_after: []
+  retained_edge_decisions: []
+  removed_edge_decisions:
+  - source: Intraspecific Metabolic Competition
+    target: Rhizosphere-Endosphere Niche Partitioning
+    rationale: Resource competition is an interpretation of associated traits, not a selectively demonstrated driver
+      of compartment partitioning.
+  - source: L-Fucose Pathway Colonization Enhancement
+    target: Rhizosphere-Endosphere Niche Partitioning
+    rationale: The separate mutant root-recovery endpoint does not establish community compartment mediation.
+  references:
+  - PMID:41543249
+  issues:
+  - 1471
+  allowed_changed_fields:
+  - curation_history
+  - discussions
+  - ecological_interactions
+  curation_events_added: 1
+  history_files:
+  - history/records/GLBRC_Populus_Variovorax_SynCom28/2026-10-06T113342Z-codex-0ba641.yaml
+  source_review:
+    PMID:41543249:
+      cache: references_cache/PMID_41543249.md
+      sha256: 8efa077f04918a0c73cc55cb4dcb14326c01c2a62b6a41ef727f340b19e40ec6
+      url: https://doi.org/10.1128/msystems.01605-25
+      scope: Complete cached primary main body, Methods and embedded captions read despite stale abstract_only front
+        matter; no figure images, separate supplements, raw data or analysis code independently inspected. Public
+        PMC browser path presented reCAPTCHA; no bypass.

reports/causal_graph_review/inventory.tsv

@@ -176,7 +176,7 @@ kb/communities/FourSpecies_Aerobic_Lung_Microbiome_Model.yaml	CommunityMech:0004
 kb/communities/Fucoidan_Seven_Degrader_Combinatorial_SynCom.yaml	CommunityMech:000449	61c3e21e1a830b7d0b3e2d3ee2fc7bc17ecc1724aa79414c106e04312c63f468	2	0	60.0	reports/causal_graph_review/decisions/20261006-three-records-batch42.yaml	needs_research
 kb/communities/GENIA_NineMember_MultiPollutant_Bioremediation_SynCom.yaml	CommunityMech:000352	72fcc1720a4bb6b3a0a464f98524ab2bbf999b2b73eaf71efa0fa98be433b9c3	3	1	78.3	reports/causal_graph_review/decisions/20261006-three-records-batch42.yaml	reviewed
 kb/communities/GLBRC_Exometabolite_Transwell_SynCom.yaml	CommunityMech:000091	8d1ef9c4d9c8b8b0f2c203a984747b02cfa32e10fe648d3fd5bf0a9548c8208f	2	0	70.0	reports/causal_graph_review/decisions/20261006-three-records-batch42.yaml	needs_research
-kb/communities/GLBRC_Populus_Variovorax_SynCom28.yaml	CommunityMech:000028	e6b9446f0f493eb1c8b2666dae051146e4375795c7d5bf51f485d09cb1e8170d	3	2	100.0		pending
+kb/communities/GLBRC_Populus_Variovorax_SynCom28.yaml	CommunityMech:000028	c251f6d716591cd391dbb09ad961f6ccb7b123988b9c677d925356b8463d2c28	3	0	61.7	reports/causal_graph_review/decisions/20261006-variovorax-batch43.yaml	reviewed
 kb/communities/GLBRC_UFMP_Fermentation_Community.yaml	CommunityMech:000029	da03a0180a9a75c6105755fe0f061f05964c258bf108a21ecd7ad11c4b14b755	3	2	100.0		pending
 kb/communities/GOM_Oil_Degrading_Consortium.yaml	CommunityMech:000030	6b90550db3bc1cc1a75459278ec2541043eab31f177007facca1d41229489f2f	3	2	100.0		pending
 kb/communities/Garlic_Pseudomonas_SynCom6.yaml	CommunityMech:000095	1336853814eb3ca22d64ec0a5c8663861c7ab544704d9216731b6b60816aef76	3	0	75.0		pending

reports/causal_graph_review/summary.json

@@ -3,13 +3,13 @@
   "semantic_review_rule": "Structural checks and readiness scores do not establish causal support. Completed reviews must match the current record SHA-256, cover every node once with a rationale, and reproduce the current edge list.",
   "records": 460,
   "nodes": 1468,
-  "edges": 686,
+  "edges": 684,
   "records_without_nodes": 8,
-  "records_without_edges": 146,
+  "records_without_edges": 147,
   "review_status_counts": {
     "needs_research": 12,
-    "pending": 285,
-    "reviewed": 163
+    "pending": 284,
+    "reviewed": 164
   },
   "structural_defects": [],
   "network_audit": [