科研 Skill
参考文献真实性核验
verify-refs
Audits every citation against PubMed and CrossRef to flag fabricated or mismatched entries. Use when checking that a manuscript’s references are real before submission.
SKILL.md
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
- Manuscript or bibliography:
.md,.docx,.bib,.txt, or.tsv. - Optional project root. Default: current working directory.
- 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 "scripts/verify_refs.py" manuscript/manuscript.md --project-root .
For hooks or quick manual runs, use the wrapper:
"scripts/verify_cli.sh" manuscript/manuscript.md --offline
Manual pre-submission strict run:
"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.statusis 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).
countsandduplicate_findings[](Gate 5).submission_safe: noFABRICATED, noMISMATCH, andduplicate_findingsempty.fully_verifiedadditionally requires noUNVERIFIED.submission_safetoleratesUNVERIFIEDrows 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 anUNVERIFIEDreference as if it were checked.source_sha256andaudited_ref_ids: what was audited. A later reader compares them with the current bib; a changed bib makes the audit stale.
Workflow
- Identify the input file and project root.
- Run
scripts/verify_refs.py. - Read
qc/reference_audit.json. Every verdict comes from it; never mark a rowOKyourself. - Report all
FABRICATEDandMISMATCHrows first (fromrecords[]). - Report all
duplicate_findings[]entries (verbatim PMID/DOI duplicates — cite renumbering required). - If
UNVERIFIEDrows remain, list them as manual checks and do not call the manuscript fully submission-safe. Checknoteon every row, whatever its status:pagination_placeholder(e000–e000/in press/TBD/forthcoming) needs the citation resolved before submission;/self-reviewPhase 2.5c decides whether any is a P0 blocker. - 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
UNVERIFIEDreferences. - 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 becomesMISMATCH: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 isOKonly when the matched record's esummary authors were compared with the cited ones and agree; otherwise it staysUNVERIFIED. A declared truncation — BibTeXand 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 == truerequires the list to be empty. - Gate 6: a reference whose raw entry still carries
e000–e000,in press,TBD, orforthcomingis markedUNVERIFIEDwithnote = "pagination_placeholder". verify-refs is manuscript-agnostic and only flags;/self-reviewPhase 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 "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 "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.
示例提示词
帮我核对一下这篇稿件里的参考文献是不是都真实存在。
所属插件
安装该插件会一并获得同领域的 58 个 Skill / MCP。
同插件的其他科研 Skill
- 医学AI论文检索优化Optimizes a medical AI paper’s title, abstract and metadata for AI search engines. Use when a paper should be findable and citable by RAG tools and AI search.
- 临床数据统计分析与建模Generates reproducible Python or R code for clinical statistics and tables. Use when collected data needs Table 1, regression, survival or agreement analysis.
- 作者发表记录与策略分析Analyzes a PubMed author record to classify study types and authorship patterns. Use when planning a publication strategy from a researcher’s track record.
- 批量队列暴露结局分析Generates one cohort-analysis script per exposure-outcome pair from one template. Use when a validated method must run across many variable combinations at once.
- 样本量估算与检验效能Walks a decision tree to the right test and computes sample size with IRB text. Use when planning how many patients a study needs before data collection.
- 因果推断方法设计Runs DAG-based, instrumental-variable, difference-in-differences or RD analyses. Use when an association in observational data must support a causal claim.
- 报告规范逐条核查Audits a manuscript item by item against 49 reporting and risk-of-bias checklists. Use when checking a draft against STROBE, CONSORT, TRIPOD or similar guidelines.
- 临床数据清洗与画像Profiles a clinical dataset and generates cleaning code in staged, approved steps. Use when a CSV or Excel dataset needs missing-value and outlier checks before analysis.
- 跨国队列数据协调对比Harmonizes variables and runs parallel weighted analyses across national surveys like KNHANES and NHANES. Use when comparing one exposure-outcome link across countries.
- 变量定义与截断值溯源Maps exposure, outcome and covariate definitions to a cited guideline and DB columns. Use when variable cutoffs need a citable basis before the study protocol.
- 本地化临床数据脱敏Runs a local, network-free script that detects and removes identifiers from clinical data. Use when data may contain PHI and must be de-identified before LLM analysis.
- AI对标专家基准研究设计Plans the arms, rubrics, calibration probes and reviewer panel for an AI-vs-expert benchmark. Use before data collection when designing an AI-vs-human rating study.
- 放射影像AI研究设计审查Reviews a study’s analysis unit, cohort logic, leakage risk and reporting-guideline fit. Use when checking a clinical or medical-AI study design before drafting.
- 传染病仓室模型拟合Builds and fits SIR, SEIR and SEIHR compartmental models with Rt estimation. Use when an outbreak or epidemic’s spread needs mathematical modeling in Python.
- ICMJE利益冲突声明填写Clones a pre-filled synthetic ICMJE COI form, replacing only date, name and title. Use when each author on a submission needs a disclosure form filled quickly.
- 机构Word表单内容填充Fills an institutional Word form, such as an IRB protocol, without breaking its layout. Use when drafted content must be rendered into a CJK-aware .docx template.
- 队列数据库选题挖掘Profiles a longitudinal cohort database and ranks research-topic proposals with gap evidence. Use when looking for a question an available cohort or registry can answer.
- 投稿期刊匹配推荐Matches a manuscript abstract against curated journal profiles and ranks candidates. Use when choosing where to submit, or re-targeting after a rejection.
- DOI全文批量获取Batch-downloads open-access PDFs for a list of DOIs via Unpaywall, PMC and OpenAlex. Use when full texts are needed for a screening set such as a meta-analysis.
- 数据集编码本自动生成Profiles every variable in a tabular dataset into a structured, citable codebook. Use when a CSV, Excel or Stata dataset needs a data dictionary before analysis.
- 全球卫生数据获取分析Accesses and analyzes global health data from WHO GHO, IHME GBD and Our World in Data. Use when DALYs, age-standardized rates or choropleth maps are needed.
- 医学AI项目基金申请书撰写Structures a grant proposal’s significance, innovation, approach, aims and milestones. Use when drafting a research funding proposal, including a Korean grant plan.
- 卫生经济学评价建模Builds Markov cohort models with QALYs, PSA and cost-effectiveness planes. Use when a health intervention needs a formal cost-effectiveness evaluation.
- AI写作痕迹识别改写Scans a manuscript for AI writing patterns and rewrites flagged passages carefully. Use when a draft or response letter reads as obviously AI-written.
常见问题
参考文献真实性核验能做什么?
Audits every citation against PubMed and CrossRef to flag fabricated or mismatched entries. Use when checking that a manuscript’s references are real before submission。
如何安装参考文献真实性核验?
在科搭 Codar 左侧点击「插件」,在「组件」页签搜索「参考文献真实性核验」并安装;在 Claude Code 或 Codex 中可以发送本页的安装提示词,或下载 Skill 包后解压到 ~/.claude/skills 或 ~/.codex/skills。
参考文献真实性核验需要哪些权限?
该科研 Skill没有声明额外权限;执行具体操作时仍由 Agent 按自身规则请求确认。
参考文献真实性核验经过官方审核吗?
是。SpeedSkills 收录的每个版本都会记录内容摘要与权限声明,当前版本已通过安全审核。