# Adversarial review log — infection-triggered-autoimmunity

**Authoring model:** Sol 6.1. **Adversarial reviewer:** Claude Opus 5.5 (Anthropic). **Date:** 2026-10-14.
**Round:** 2 (2026-10-14). **Canonical path:** `consensus_consistency_audit_2026-10-09/revised/infection-triggered-autoimmunity/infection-triggered-autoimmunity.review.md`.

## Summary

| Level before | Level after | Confidence before | Confidence after | Critical | Major | Minor | Editorial | Approval |
|---|---|---|---|---|---|---|---|---|
| Β — Associational | Β — Associational (unchanged) | Moderate | Moderate (re-derived under lead policy; band stands) | 0 | 2 | 1 | 0 | Approved as revised |

PMIDs checked 70/70, all resolved to the stated first author, title, journal and DOI. Retraction / expression-of-concern
screen: none found. Three linked errata verified (PMID 33055421 → Pender [18]; PMID 30455315 → Zhao [52];
PMID 29724927 → Manfredo Vieira [61]). Load-bearing claims spot-checked: 12. Validator (`validate_review.py` v2.3,
HSEF v5.3 tree): **pass** — 0 FAIL, 1 WARN (`v5.claim_form` "phenomenon", rejected per standing convention), 31 PASS.

## Findings

| ID | Severity | Location | Original → corrected | Evidence |
|---|---|---|---|---|
| IA-1 | Major | Consensus provenance reconciliation block | "The earlier adversarial approval refers to its original edition" → "No earlier adversarial approval exists for any edition of this review" | No edition carries an approval section or approval line; the 2026-10-06 pass covered 30 other slugs. Spliced by the Oct 9 revision. |
| IA-2 | Minor | REFERENCES entry 41 (Sakaguchi 1995, J Immunol) | "DOI: ." → "DOI: none assigned (pre-DOI Journal of Immunology record)." | Known empty-DOI renderer defect. PubMed holds no DOI for PMID 7636184, so the repair is an explicit statement rather than a fabricated identifier. |
| IA-3 | Major | AT A GLANCE (new **Confidence basis** paragraph); `final_score.confidence_rationale` and its two mirrored copies | "density, heterogeneous controls and **incomplete access** set the confidence" → rationale grounded in evidence density and reproducibility, access depth excluded | Lead policy; `hsef-framework.md` § Confidence band. Two-step test: access depth was not the only basis — induced/engineered systems, sign-flipping microbial results, a nonrandomized abatacept cohort and partly audited platform overlap are valid determinants — so **Moderate stands**. |

## Rejected findings (checked, no change)

1. **The overall Β.** Argued against, then upheld. SC1 is a descriptive antecedence row, not a causal conjunct, so a
   weakest-link conjunction including it drives the headline below the weakest causal subclaim; the review names this
   an aggregation artifact. Retention is correct because the rule was pre-registered before scoring, SKILL.md Step 3
   exists to prevent exactly this kind of post-hoc rule-shopping, and both alternatives (primary-subclaim Ε,
   scope-matched Ε) are published in four places in the document.
2. **AT A GLANCE headline reading stronger than Β.** It reports what the matrix contains, with the aggregation basis
   stated in the adjacent paragraph. Accurate.
3. **"Best-supported rival" renderer defect.** Rank 1 (gene-environment threshold, M5) is `is_nominal: false`; the
   nominal M1 ranks 3. Defect absent.
4. **Six print-vs-e-pub year differences of one** (Kim, Kamitani, Traka, McCall, Lindfors, Zegarra-Ruiz). Standing
   convention.
5. **Three apparent title mismatches** — PubMed rendering "CD4+ T cells"/"CD8+ T cells" as "CD4T"/"CD8T", and
   dropping the italicised species name in the Mukherjee title. The review's text is correct; the PubMed field is
   lossy. Standing convention.
6. **`v5.claim_form` "phenomenon" warning.** Standing convention.
7. **REVIEWER NOTES vs spliced Consensus section.** No contradiction; the notes record the Consensus work performed.
8. **Momentum.** Already "Mechanistic only"; the named artifacts are completed publications and a completed phase I,
   which cannot support "Decisive work underway".
9. **Collective authors.** None in this reference list; no leading-comma defect.

## Residual limitations

Access depth unchanged: 6 partial-body and 64 abstract/status examinations, no complete-paper or supplement audit.
The three linked errata remain unexamined in substance. Partial-body-sourced detail for Godschalk, Lanz, Wen, Krogvold
and Younis could not be re-verified beyond abstracts and is not contradicted. Cohort and isolate-platform overlap
remains only partly audited.

## Changed files

| File | sha256 | bytes |
|---|---|---|
| infection-triggered-autoimmunity.review.md | 8590f4e2b7d5349e09eb457b26c4267eff1b23d122d82ce0227b44611ec18be3 | 120721 |
| infection-triggered-autoimmunity.scorecard.json | 686349b36af9ff83321daefce68cb3182fe107d9b13948373866cc16927da4c1 | 352350 |

`SHA256.txt` regenerated; `validation-report.json` re-run. Pre-edit copies in `sol61-original/`.


## Collective-author screen (both known forms)

Screened for a dropped author leaving a stray leading comma, for a consortium name truncated to its last word so
that it reads as an ordinary surname ("Group et al.", "Network et al.", "Consortium et al."), and for dropped
nobiliary particles. Method: every reference's rendered first author compared against the PubMed record,
including records whose first author object comes back empty because the article is credited to a consortium.
Result: no collective first author in this list; particles intact ("Gonzalez-Figueroa", "Zegarra-Ruiz"). Five records carry a consortium mid-list, which does not affect the rendered form. No defect.

## Line-ending note

The edit pass initially normalised these files from CRLF to LF. That was unintended and has been reversed: the
review and score card are back on CRLF, matching their pre-edit state and the rest of the package, and the hashes
below are the final CRLF files.

**Approval:** Following adversarial review and the corrections listed above, the revised document is approved. — Claude Opus 5.5 (Anthropic), 2026-10-14.