Verify References (Audit-Only)

Audit an existing manuscript or bibliography for fabricated or mismatched references. This skill never writes to references/ or manuscript/_src/refs.bib: corrections flow through /lit-sync (Zotero → Better BibTeX → refs.bib), and /manage-project's scripts/validate_project_contract.py flags any references/* file written here as drift. It does not discover literature (/search-lit), and it never replaces a missing or bad citation with a plausible alternative — replacements go through /search-lit or the user.

Inputs

1. Manuscript or bibliography: .md, .docx, .bib, .txt, or .tsv.
2. Optional project root. Default: current working directory.
3. Optional flags: --offline (extract and classify without API calls), --timeout N (HTTP timeout seconds), --strict, --no-openalex.

For markdown manuscripts with pandoc [@bibkey] citations, run a citation-key check first (/manage-refs's check_citation_keys.py, or your reference manager's). It catches mis-keyed cites; verify_refs.py catches fabricated metadata.

Deterministic Script

Run the bundled script rather than verifying citations by memory:

python "${CLAUDE_SKILL_DIR}/scripts/verify_refs.py" manuscript/manuscript.md --project-root .

For hooks or quick manual runs, use the wrapper:

"${CLAUDE_SKILL_DIR}/scripts/verify_cli.sh" manuscript/manuscript.md --offline

Manual pre-submission strict run:

"${CLAUDE_SKILL_DIR}/scripts/verify_cli.sh" manuscript/index.qmd --strict

--strict forbids --offline and exits non-zero on any UNVERIFIED row. Read references/manual_checkpoint_guide.md for when to run it and what to do per status.

Lookups run PubMed (PMID) → CrossRef (DOI; doi.org's handle registry when CrossRef answers 404) → OpenAlex, with a PubMed title search last. Every OK row rests on DOI, PMID, CrossRef, or PubMed title evidence; a failed lookup is recorded as UNVERIFIED, never silently passed. A title search, in either index, counts only when a returned title matches the cited one (the same token-similarity guard), so a made-up title cannot earn OK from a search that merely returned something.

OpenAlex tier. It recovers conference and non-DOI citations (NeurIPS / ICLR / ACL, common in medical-AI manuscripts) that PubMed and CrossRef miss, and is called only when no author list was obtained yet. It resolves by DOI, otherwise by a title search behind a token-similarity guard so a fabricated title cannot earn OK. OpenAlex names have no family/given split, so they support only an existence check and a tolerant first-author membership check — never the strict positional or author-count MISMATCH, which stays with PubMed efetch / CrossRef. An OpenAlex miss is UNVERIFIED, never FABRICATED. --no-openalex restricts verification to PubMed + CrossRef.

Output Contract (v1.3.0)

qc/reference_audit.json (schema_version 4) is the only output; this skill is its sole writer and writes no TSV and no library.bib.

• records[]: per reference, status, note, evidence, cited_authors[] / actual_authors[], cited_author_count / actual_author_count. status is one of:
  • OK: a lookup confirmed the work (DOI, PMID, or a title search passing the similarity guard) and the authors agree.
  • MISMATCH: the work exists but the citation is wrong: its authors differ (Gate 4), or its DOI/PMID does not exist while a lookup still found the work (note = "wrong identifier…").
  • FABRICATED: the cited identifier does not exist and nothing else found the work (details below).
  • UNVERIFIED: nothing confirmed or refuted it (a failed lookup, a DOI held only by another registration agency that no index confirmed, no identifier and no title match, a PubMed title-only match whose authors could not be compared or differ (note = "title_only: …"), or --offline).
• counts and duplicate_findings[] (Gate 5).
• submission_safe: no FABRICATED, no MISMATCH, and duplicate_findings empty. fully_verified additionally requires no UNVERIFIED.
• submission_safe tolerates UNVERIFIED rows because offline runs produce them; before a submission, resolve each one (confirm it by hand and say how, or remove the citation) — do not ship an UNVERIFIED reference as if it were checked.
• source_sha256 and audited_ref_ids: what was audited. A later reader compares them with the current bib; a changed bib makes the audit stale.

Workflow

1. Identify the input file and project root.
2. Run scripts/verify_refs.py.
3. Read qc/reference_audit.json. Every verdict comes from it; never mark a row OK yourself.
4. Report all FABRICATED and MISMATCH rows first (from records[]).
5. Report all duplicate_findings[] entries (verbatim PMID/DOI duplicates — cite renumbering required).
6. If UNVERIFIED rows remain, list them as manual checks and do not call the manuscript fully submission-safe. Check note on every row, whatever its status: pagination_placeholder (e000–e000 / in press / TBD / forthcoming) needs the citation resolved before submission; /self-review Phase 2.5c decides whether any is a P0 blocker.
7. If the user needs a human-readable table, summarize from records[] in chat — do not write a TSV.

Quality Gates

• Gate 1: stop submission if any row is FABRICATED.
• Gate 2: require user confirmation before accepting UNVERIFIED references.
• Gate 3: rerun after any reference edits.
• Gate 4 (full-author cross-check): the authoritative author list comes from PubMed efetch.fcgi (XML) when a PMID is present — preferred because CrossRef is unreliable for given names — with CrossRef and PubMed esummary as fallbacks. For BibTeX inputs every cited family name is compared index-by-index, and the cited and source author counts are compared, after normalizing case, diacritics, hyphen vs space, and name particles ("von", "van", "de"); one name may be a whole word of the other ("Garcia" / "Garcia-Lopez"), never a mere substring ("Li" / "Williams"). A row whose DOI/PMID resolves but whose authors differ at any index or in count becomes MISMATCH: note = "first-author hallucination suspected" for the first author, note = "non-first-author hallucination or count mismatch" for #2..#N or the count. A correct first author does not establish the rest of the list. Plain-text / TSV inputs, and lists that cannot be parsed confidently, degrade to the first-author check (skipped if even that is empty). A PubMed title-only match is OK only when the matched record's esummary authors were compared with the cited ones and agree; otherwise it stays UNVERIFIED. A declared truncation — BibTeX and others, or a _audit_truncated = <N> field — turns a shorter-than-source count into a note; citing more authors than the source is always flagged.
• Gate 5: verbatim PMID or normalized-DOI duplicates in the reference list are MAJOR findings in duplicate_findings[]; submission_safe == true requires the list to be empty.
• Gate 6: a reference whose raw entry still carries e000–e000, in press, TBD, or forthcoming is marked UNVERIFIED with note = "pagination_placeholder". verify-refs is manuscript-agnostic and only flags; /self-review Phase 2.5c, with the manuscript in hand, decides whether the citation is method- or headline-load-bearing and hence a P0 blocker.

Citation-metadata confusion is not fabrication. DOI-suffix digits that look like, but differ from, the article number (a DOI tail "77196" against article 26068) are cosmetic. When the identifier resolves and the authors match, do not report the row as fabricated: the script returns FABRICATED only for an identifier that does not exist — a PMID PubMed has no record of, or a DOI that CrossRef answers 404 for and that doi.org's handle registry (which spans every registration agency: CrossRef, DataCite, mEDRA, JaLC) reports as not found — and only when no lookup found the work some other way. After a CrossRef 404 the DOI verdict is:

doi.org handle registry	another lookup finds the work	status
not found	no	FABRICATED
not found	yes (a title search passing the similarity guard)	MISMATCH — a real paper cited with a wrong DOI
found (registered with another agency)	no / yes	UNVERIFIED / OK
lookup failed (network, 5xx, unexpected reply)	no / yes	UNVERIFIED / OK

A DOI field that is not a well-formed DOI (a placeholder such as n/a) is not sent to doi.org and stays UNVERIFIED, and so does a legacy SICI DOI (10.1002/(SICI)…) that doi.org does not find, since its punctuation is easily cut in extraction. A real identifier with wrong authors is MISMATCH (Gate 4).

Claim Fidelity — does the source say what you say it says?

verify_refs.py answers whether a reference is real and whose it is, not whether the sentence citing it is true of it. scripts/check_claim_fidelity.py checks the claims that have a checkable answer against full texts already downloaded and converted (/fulltext-retrieval produces exactly that layout; this script never fetches anything):

python3 "${CLAUDE_SKILL_DIR}/scripts/check_claim_fidelity.py" \
  --manuscript manuscript/manuscript.md \
  --fulltext-dir fulltext/ --bib manuscript/_src/refs.bib \
  --out qc/claim_fidelity.json --strict

Verdict	Severity	Fires when
CITED_QUOTE_ABSENT	major	Quoted text attributed to a source is not in it in any reading order.
CITED_QUOTE_UNRESOLVED	prompt	The quote matched with a word or two missing — the signature of a dirty extraction, not of a fabrication. Look; do not assume.
ATTRIBUTION_UNSUPPORTED	prompt	Not one content word of the attributed claim appears in the source, in any form. Paraphrase normally keeps at least one of the source's own terms.
ORDINAL_CLAIM_UNSUPPORTED	prompt	"reports three strategies [12]" where the source discusses that noun but never that count near it.

Only the quote verdict can fail --strict; the others are prompts to go read the source, because paraphrase is legitimate. Read the "not checked" lines: a citation with no full text on disk is reported unresolved, and a source whose extracted text is an abstract is reported as too short to judge. Silence means "nothing checkable was wrong", not "everything is right".

Known limits: a quote whose words all appear in order with source tokens between them (INTERLEAVED) is treated as verified, so a quotation that drops a source word such as "not" produces no finding; reread every quotation that changes a negation or qualifier.

Sentence-level source evidence table

qc/claim_fidelity.json also carries evidence_rows: one per recognized prose sentence/citation pair, with manuscript coordinates, source-text and PDF hashes, advisory retrieval identity, and an assessor-entered comparison that starts not_assessed even when the bibliographic status is OK and no probe fires.

python3 "${CLAUDE_SKILL_DIR}/scripts/check_claim_fidelity.py" \
  --manuscript manuscript/manuscript.md --bib manuscript/_src/refs.bib \
  --fulltext-dir fulltext/ --retrieval-report pdfs/retrieval_report.json \
  --reference-audit qc/reference_audit.json \
  --out qc/claim_fidelity.json --evidence-table qc/claim_fidelity.md

Enter pages, excerpts, metric/unit/denominator, population, direction, and a named assessment only after inspecting the actual source; equal numbers or matching words do not establish support. Record whether the assessor used AI assistance, and never describe an AI-generated assessment as human approval. Rerun with --reviewed-report qc/claim_fidelity.json to retain annotations; changed inputs leave old assessments unresolved. The Markdown table is a derived view, not a second editable evidence store. Read references/claim_evidence_workflow.md before entering or carrying forward an assessment.