Preserve source provenance in snippet cache audits (batch 26) by realmarcin · Pull Request #1363 · CultureBotAI/CommunityMech

realmarcin · 2026-10-06T04:05:52Z

Scope

Fix independent snippet-cache provenance. No biological record, cache file, history, taxonomy, node, or arrow is edited. Corpus review coverage remains 123 reviewed / 334 pending / 3 needs_research.

Closes #1362. Source-coverage and literal-quotation follow-up #596 remains open.

Repair And Review

Explicit selected-excerpt appendices are excluded before format-specific cache trust. Retrieved abstract prefixes and later independently retrieved full text remain available. Empty/marker-only text is not certified. The existing Markdown selected_excerpts guard is unchanged.

Self-adversarial review caught a candidate-fix defect: shortened, indented or differently cased pseudo-retrieval headers could end exclusion too early. The final boundary reuses the exact established FULLTEXT_MARKER expression. Initial 14 failing tests and these three later adversarial failures are preserved as historical evidence; all 22 new regression cases now pass.

Verification

60 passed in 131.45s (0:02:11); 324 passed in 188.25s (0:03:08); 460 records and 920 cache files byte-preserved; 641 of 643 cited-reference views unchanged.
Graph ranking and all 25 existing network warnings are unchanged; no new structural errors. Actual post-commit just check-docs-current passed and left the worktree clean. The raw warning-only network exits are retained, not relabeled zero-finding passes.

The two affected records move from 43 MATCH to 22 MATCH / 19 MISMATCH / 1 ASSEMBLED / 1 RENDERING. DCPP's 19 mismatches are gaps against its independently retrieved abstract, not proof that its biology is false. RAMC keeps its retrieved full text and exposes one assembled medium quote plus one rendering difference. No selected text was copied into trusted source content to hide these findings. The upstream reference-validator itself is not modified or asserted to enforce this guard.

Committed output: cache-provenance-20261006-batch26.json and adversarial-review-20261006-batch26.yaml under reports/causal_graph_review/. This is self-review, not independent approval or a new primary-paper review.

Stacking

Draft dependency on #1361. Do not merge into the topic parent. Require exact-head CI and the protected main queue after parents merge; move dependents before deleting branches.

realmarcin

Code-based self-adversarial review at exact head d35e526e30385056f3b99c9a93c85a0d791b309b, not independent approval.

Addressed #1362. The original selected-excerpt trust defect and the malformed-boundary defect discovered in the first candidate are corrected with 22 regression cases. All 643 cited-reference cache views were compared and only the two intended views changed. 60 passed in 131.45s (0:02:11); 324 passed in 188.25s (0:03:08); 460 records and 920 cache files byte-preserved; 641 of 643 cited-reference views unchanged.

The shared matcher thresholds are unchanged; selected sections cannot certify themselves, while actual retrieved prefixes and later full-text markers survive. All records/caches, the graph inventory, and the existing network warnings are preserved. Source coverage gaps are not counted as passing matches: DCPP has 19 MISMATCH items; RAMC has one ASSEMBLED and one RENDERING item alongside its 20 exact matches. #596 remains open, and the upstream reference-validator is not represented as fixed.

Committed validation/self-review output retains both rounds of expected pre-fix failures and the complete final rerun, including exact tested-code hashes. Actual post-commit docs-current passed cleanly. This repair adds no biological record reviews or graph structure, spends no Edison credits, and does not certify the full remaining corpus. Protected exact-head CI and main queue remain mandatory; no topic-parent merge or premature branch deletion.

github-actions · 2026-10-06T04:11:23Z

Network integrity findings

Warnings only — a member with no interaction yet, or a participant matched by ontology id rather than by name, or one on a community-level interaction that resolves to no member. Reported, but does not fail the build.

Network Integrity Audit Report
================================================================================

0 error, 25 warning across 10 records with findings
Only error-severity findings fail the build.

Algal_Methanotroph_Biogas_Valorization_Coculture
--------------------------------------------------------------------------------
  [warning] DISCONNECTED: Taxon 'Leptolyngbya' has no interactions

Total: 1 issues (0 error, 1 warning)

At_RSPHERE_SynCom
--------------------------------------------------------------------------------
  [warning] DISCONNECTED: Taxon 'Flavobacterium sp.' has no interactions
  [warning] DISCONNECTED: Taxon 'Pseudomonas sp.' has no interactions
  [warning] DISCONNECTED: Taxon 'Rhodococcus sp.' has no interactions
  [warning] DISCONNECTED: Taxon 'Streptomyces sp.' has no interactions
  [warning] DISCONNECTED: Taxon 'Variovorax sp.' has no interactions

Total: 5 issues (0 error, 5 warning)

Australian_Lead_Zinc_Polymetallic
--------------------------------------------------------------------------------
  [warning] DISCONNECTED: Taxon 'Acidithiobacillus ferrooxidans' has no interactions
  [warning] DISCONNECTED: Taxon 'Leptospirillum ferriphilum' has no interactions

Total: 2 issues (0 error, 2 warning)

Bacillus_G12_Y4_X25_Tobacco_Biocontrol_SynCom
--------------------------------------------------------------------------------
  [warning] UNKNOWN_TARGET: Target taxon 'tobacco' not found in taxonomy section (community-level scope)
  [warning] UNKNOWN_TARGET: Target taxon 'tobacco' not found in taxonomy section (community-level scope)

Total: 2 issues (0 error, 2 warning)

Bacillus_siamensis_vallismortis_HT_Masson_Pine_SynCom
--------------------------------------------------------------------------------
  [warning] UNKNOWN_TARGET: Target taxon 'Fusarium oxysporum' not found in taxonomy section (community-level scope)
  [warning] UNKNOWN_TARGET: Target taxon 'Pinus massoniana' not found in taxonomy section (community-level scope)
  [warning] UNKNOWN_TARGET: Target taxon 'Pinus massoniana' not found in taxonomy section (community-level scope)

Total: 3 issues (0 error, 3 warning)

Brocadia_Low_Nitrogen_Anammox_Consortium
--------------------------------------------------------------------------------
  [warning] DISCONNECTED: Taxon 'Ca. Kuenenia early-acclimation lineage' has no interactions

Total: 1 issues (0 error, 1 warning)

Mangrove_Benzene_MFC_Bioanode_Consortium
--------------------------------------------------------------------------------
  [warning] DISCONNECTED: Taxon 'Pelobacter' has no interactions
  [warning] DISCONNECTED: Taxon 'Pseudomonas' has no interactions

Total: 2 issues (0 error, 2 warning)

Nitrifying_Wastewater_Ammonia_Oxidizing_Consortium
--------------------------------------------------------------------------------
  [warning] DISCONNECTED: Taxon 'Legionella MAGs' has no interactions

Total: 1 issues (0 error, 1 warning)

Sclerotinia_Sclerotia_12Strain_Biocontrol_SynCom
--------------------------------------------------------------------------------
  [warning] UNKNOWN_TARGET: Target taxon 'Sclerotinia sclerotiorum' not found in taxonomy section (community-level scope)
  [warning] UNKNOWN_TARGET: Target taxon 'Sclerotinia sclerotiorum' not found in taxonomy section (community-level scope)

Total: 2 issues (0 error, 2 warning)

Space_Habitat_SevenMember_Stress_Tolerance_SynCom
--------------------------------------------------------------------------------
  [warning] DISCONNECTED: Taxon 'Acinetobacter radioresistens 50V1' has no interactions
  [warning] DISCONNECTED: Taxon 'Micrococcus luteus W4-Al-K1' has no interactions
  [warning] DISCONNECTED: Taxon 'Propionibacterium cyclohexanicum TA-12T' has no interactions
  [warning] DISCONNECTED: Taxon 'Pseudomonas antarctica CMS 35T' has no interactions
  [warning] DISCONNECTED: Taxon 'Staphylococcus capitis K1-2-2-23' has no interactions
  [warning] DISCONNECTED: Taxon 'Streptococcus halotolerans HTS9T' has no interactions

Total: 6 issues (0 error, 6 warning)

The full report is attached to the workflow run as an artifact.

realmarcin

Self-adversarial review refreshed at exact head 39008cba4bd3858e6fee7a0a9d36bd4665a9bafc
after protected main merge of #1307. This is not independent approval.

The entire tree remains identical to the preceding reviewed tree 2a18568fda1ef8544ba9fbff1ec72425cf785169.
The batch-owned binary patch is byte-identical
(SHA-256 8b3f71f8a6dcab461480539f08b038e05c6e64148e33aac36bdb8c70ecc03050); only ancestry changed.
Scientific records, caches, source-review output, histories, regression tests, and generated pages
are unchanged. Historical validation results and recorded source/access limitations therefore
remain applicable, but required CI must pass at this new head. No new source review or local test
rerun is claimed by this ancestry-only update.

The protected main queue remains mandatory. Keep dependents on their reviewed parent until that
parent is confirmed merged; move dependents before deleting merged branches.

realmarcin

Self-adversarial review refreshed at exact head 4cef0dc8eae33054a5dc0330bdecf1112dac357d
after protected main merge of #1312. This is not independent approval.

The entire tree remains identical to the preceding reviewed tree 2a18568fda1ef8544ba9fbff1ec72425cf785169.
The batch-owned binary patch is byte-identical
(SHA-256 8b3f71f8a6dcab461480539f08b038e05c6e64148e33aac36bdb8c70ecc03050); only ancestry changed.
Scientific records, caches, source-review output, histories, regression tests, and generated pages
are unchanged. Historical validation results and recorded source/access limitations therefore
remain applicable, but required CI must pass at this new head. No new source review or local test
rerun is claimed by this ancestry-only update.

The protected main queue remains mandatory. Keep dependents on their reviewed parent until that
parent is confirmed merged; move dependents before deleting merged branches.

realmarcin

Self-adversarial review refreshed at exact head 36ef60bde73e97c203a759875cfdde9e66f42d25
after protected main merge of #1315. This is not independent approval.

The entire tree remains identical to the preceding reviewed tree 2a18568fda1ef8544ba9fbff1ec72425cf785169.
The batch-owned binary patch is byte-identical
(SHA-256 8b3f71f8a6dcab461480539f08b038e05c6e64148e33aac36bdb8c70ecc03050); only ancestry changed.
Scientific records, caches, source-review output, histories, regression tests, and generated pages
are unchanged. Historical validation results and recorded source/access limitations therefore
remain applicable, but required CI must pass at this new head. No new source review or local test
rerun is claimed by this ancestry-only update.

The protected main queue remains mandatory. Keep dependents on their reviewed parent until that
parent is confirmed merged; move dependents before deleting merged branches.

realmarcin

Self-adversarial review refreshed at exact head cbb2ba27fe60c02bfa952525f289c5c3ed43dd89
after protected main merge of #1318. This is not independent approval.

The entire tree remains identical to the preceding reviewed tree 2a18568fda1ef8544ba9fbff1ec72425cf785169.
The batch-owned binary patch is byte-identical
(SHA-256 8b3f71f8a6dcab461480539f08b038e05c6e64148e33aac36bdb8c70ecc03050); only ancestry changed.
Scientific records, caches, source-review output, histories, regression tests, and generated pages
are unchanged. Historical validation results and recorded source/access limitations therefore
remain applicable, but required CI must pass at this new head. No new source review or local test
rerun is claimed by this ancestry-only update.

The protected main queue remains mandatory. Keep dependents on their reviewed parent until that
parent is confirmed merged; move dependents before deleting merged branches.

realmarcin

Self-adversarial review refreshed at exact head 49c5e767a9a636e1e4ab2e858834ffe00727e0b5
after protected main merge of #1323. This is not independent approval.

The entire tree remains identical to the preceding reviewed tree 2a18568fda1ef8544ba9fbff1ec72425cf785169.
The batch-owned binary patch is byte-identical
(SHA-256 8b3f71f8a6dcab461480539f08b038e05c6e64148e33aac36bdb8c70ecc03050); only ancestry changed.
Scientific records, caches, source-review output, histories, regression tests, and generated pages
are unchanged. Historical validation results and recorded source/access limitations therefore
remain applicable, but required CI must pass at this new head. No new source review or local test
rerun is claimed by this ancestry-only update.

The protected main queue remains mandatory. Keep dependents on their reviewed parent until that
parent is confirmed merged; move dependents before deleting merged branches.

realmarcin

Self-adversarial review refreshed at exact head 928904595787ec05a540f1efa7ef6ecdd6f5071f
after protected main merge of #1328. This is not independent approval.

The entire tree remains identical to the preceding reviewed tree 2a18568fda1ef8544ba9fbff1ec72425cf785169.
The batch-owned binary patch is byte-identical
(SHA-256 8b3f71f8a6dcab461480539f08b038e05c6e64148e33aac36bdb8c70ecc03050); only ancestry changed.
Scientific records, caches, source-review output, histories, regression tests, and generated pages
are unchanged. Historical validation results and recorded source/access limitations therefore
remain applicable, but required CI must pass at this new head. No new source review or local test
rerun is claimed by this ancestry-only update.

The protected main queue remains mandatory. Keep dependents on their reviewed parent until that
parent is confirmed merged; move dependents before deleting merged branches.

realmarcin

Self-adversarial review refreshed at exact head cd9d40fbb3d7b8fce1f8f0119b03a18609cabb38
after protected main merge of #1334. This is not independent approval.

The entire tree remains identical to the preceding reviewed tree 2a18568fda1ef8544ba9fbff1ec72425cf785169.
The batch-owned binary patch is byte-identical
(SHA-256 8b3f71f8a6dcab461480539f08b038e05c6e64148e33aac36bdb8c70ecc03050); only ancestry changed.
Scientific records, caches, source-review output, histories, regression tests, and generated pages
are unchanged. Historical validation results and recorded source/access limitations therefore
remain applicable, but required CI must pass at this new head. No new source review or local test
rerun is claimed by this ancestry-only update.

The protected main queue remains mandatory. Keep dependents on their reviewed parent until that
parent is confirmed merged; move dependents before deleting merged branches.

realmarcin

Self-adversarial review refreshed at exact head b087c3d196729d74c0bd60ae3db9f8fefa9bbe86
after protected main merge of #1339. This is not independent approval.

The entire tree remains identical to the preceding reviewed tree 2a18568fda1ef8544ba9fbff1ec72425cf785169.
The batch-owned binary patch is byte-identical
(SHA-256 8b3f71f8a6dcab461480539f08b038e05c6e64148e33aac36bdb8c70ecc03050); only ancestry changed.
Scientific records, caches, source-review output, histories, regression tests, and generated pages
are unchanged. Historical validation results and recorded source/access limitations therefore
remain applicable, but required CI must pass at this new head. No new source review or local test
rerun is claimed by this ancestry-only update.

The protected main queue remains mandatory. Keep dependents on their reviewed parent until that
parent is confirmed merged; move dependents before deleting merged branches.

realmarcin

Self-adversarial review refreshed at exact head a1fd521c83d1b02b3a891d48e43bd4199111f01a
after protected main merge of #1346. This is not independent approval.

The entire tree remains identical to the preceding reviewed tree 2a18568fda1ef8544ba9fbff1ec72425cf785169.
The batch-owned binary patch is byte-identical
(SHA-256 8b3f71f8a6dcab461480539f08b038e05c6e64148e33aac36bdb8c70ecc03050); only ancestry changed.
Scientific records, caches, source-review output, histories, regression tests, and generated pages
are unchanged. Historical validation results and recorded source/access limitations therefore
remain applicable, but required CI must pass at this new head. No new source review or local test
rerun is claimed by this ancestry-only update.

The protected main queue remains mandatory. Keep dependents on their reviewed parent until that
parent is confirmed merged; move dependents before deleting merged branches.