Nature Data

Nature Data Availability — Router

This skill is split into two layers:

• A static layer under static/ that holds versioned, reusable content fragments (the default stance and source hierarchy, the Chinese-user operating mode, and the workflow with output format).
• A dynamic layer (this file plus manifest.yaml) that loads the core every time and reaches for the deeper policy/repository/FAIR references only when a step needs them.

Do not try to apply the data-availability logic from memory or from this router. Always load fragments from disk as described below.

Routing protocol

Follow these four steps every time the skill is invoked.

1. Load the manifest and the core layer

Read manifest.yaml. Then read every file listed under always_load:

• static/core/stance.md — what the data-availability package is, the default stance, and the source hierarchy.
• static/core/chinese-mode.md — how to operate when the user writes in Chinese (accept Chinese, draft English, convert terms precisely).
• static/core/workflow.md — the eight-step workflow and the output format.

2. No content axis — confirm journal and language inline

Unlike nature-writing or nature-figure, nature-data has no fragment axis. Its variation is handled at runtime, not by loading different content bodies:

• journal/article type — if journal-specific instructions conflict with this skill, follow the journal.
• access route — each dataset is classified into one route (public repository, controlled access, within paper, reused public, third-party restricted, justified request, or not applicable).
• user language — if the user writes Chinese, follow core/chinese-mode.md and add the 中文核对 block.

3. Run the workflow

Follow the eight-step workflow in core/workflow.md: identify the journal, inventory every supporting dataset, classify each into one access route, choose repository and identifier strategy before drafting, draft the statement with explicit dataset-to-location mapping, add formal dataset citations, run the FAIR/metadata audit, and return ready-to-paste text plus unresolved fields.

Do not invent DOIs, accession numbers, repository names, licences, embargo dates, ethics approvals, access committees, or data-use conditions. Flag "available upon request" as weak unless there is a specific legal, ethical, commercial, or third-party restriction.

4. Reach for references only when needed

The files under references/ are deep references, not defaults. Open them on demand per the references.on_demand table in the manifest — for example references/policy-principles.md for the governing rules and edge cases, references/repository-and-identifiers.md for repository/accession/DOI choices, references/statement-patterns.md for ready-to-adapt statements, references/fair-metadata-checklist.md for the FAIR audit, references/chinese-author-alignment.md for Chinese wording, and references/source-basis.md to justify a rule with its official source.

Why this split

• The static layer is versioned and reviewable; the core stays small for a normal statement.
• The dynamic layer keeps each invocation cheap: the policy, repository, and FAIR depth load only when a step needs them.
• The router itself is short on purpose. Update fragments and references, not this file, when adding scope.
• This structure mirrors nature-writing, nature-polishing, nature-reader, nature-paper2ppt, nature-figure, nature-citation, and nature-response.

nature-data skill

A data-availability skill for preparing manuscript data statements, repository plans, dataset citations, and FAIR metadata checks in a Nature / Springer Nature publication style.

This skill is bilingual-aware. It accepts Chinese author notes covering data availability statements, data requests to the corresponding author, raw data, restricted data, or public databases, then converts them into submission-ready English with Chinese action notes for the author.

What it does

• drafts ready-to-paste Data Availability statements
• audits weak or incomplete data statements before submission
• maps each supporting dataset to a repository, accession, DOI, or access route
• distinguishes public, controlled-access, third-party, supplementary, and not-applicable cases
• prepares FAIR metadata and DataCite-style dataset citation checks
• flags missing repository records, licences, provenance, embargo details, and access conditions
• aligns Chinese author intent with Nature-style English availability wording

Source hierarchy

• Nature Portfolio and Springer Nature research data policies
• Nature Portfolio reporting standards for availability of data, code, materials, and protocols
• Scientific Data data policies for repository, rawness, preservation, and data citation practice
• FAIR Guiding Principles and DataCite metadata schema

File structure

The skill uses a router/static-dynamic split (like the other nature-* skills): a short SKILL.md router plus a manifest.yaml. nature-data is a linear workflow with no content axis, so the split is core (always loaded) plus on-demand references.

nature-data/
├── SKILL.md                     # short router
├── manifest.yaml                # always_load core + on-demand references (no axis)
├── README.md
├── agents/
│   └── openai.yaml
├── static/
│   └── core/                    # always loaded
│       ├── stance.md            # default stance + source hierarchy
│       ├── chinese-mode.md      # Chinese-user operating mode
│       └── workflow.md          # the 8-step workflow + output format
└── references/
    ├── fair-metadata-checklist.md
    ├── chinese-author-alignment.md
    ├── policy-principles.md
    ├── repository-and-identifiers.md
    ├── source-basis.md
    └── statement-patterns.md

When to use

• preparing a Data Availability statement for a Nature-family or Springer Nature journal
• deciding where to deposit data before submission
• revising "available on request" language
• handling controlled-access, human-participant, proprietary, or third-party data
• citing datasets with DOI, accession number, Handle, ARK, or repository record
• checking whether a dataset deposit is FAIR enough for publication
• converting Chinese data-availability notes into precise English submission language

Design intent

The skill should make the availability route explicit for every dataset that supports the paper's claims. It should not fabricate accessions, licences, restrictions, or repository metadata. When information is missing, it should return a usable draft plus a short list of items the author must confirm, preferably with Chinese notes when the user is working from a Chinese draft.

Chinese Author Alignment

Use this file when the user writes in Chinese, provides a Chinese Data Availability draft, or asks for bilingual wording. The goal is not to translate Chinese literally. The goal is to convert the author's Chinese description into a Nature-ready English availability route.

Core terminology

Chinese-to-English conversion rules

• Convert "本文所有数据均包含在正文和补充材料中" to a specific claim:

name Source Data files, Supplementary Tables, or repository records. If raw data are absent, say so as a risk flag rather than pretending they are included.

• Convert "可向通讯作者合理索取" only after adding:

why public sharing is impossible, who reviews requests, eligible requesters, required approvals or data-use agreement, and expected access route.

• Convert "数据因隐私原因不可公开" into a controlled-access pattern:

state privacy/consent/legal basis, public metadata if available, access committee or institution, and conditions.

• Convert "商业数据/企业数据不可公开" into a third-party or commercial restriction pattern:

name the provider or owner, request route, and whether derived or aggregate data can be shared.

• Convert "数据将在接收后上传" into an action item:

deposit before submission or create a private reviewer link if the repository supports it.

• Convert "使用公开数据集" into a citation requirement:

include source, version/release/date accessed when relevant, and dataset citation.

Bilingual intake questions

Ask only what is needed for the statement.

请确认这些字段：
1. 哪些数据支撑主文图、补充图和统计分析？
2. 每类数据是否已有仓库、DOI、登录号或审稿人私密链接？
3. 是否包含人类参与者、隐私、商业、第三方授权或国家/机构限制？
4. 如果数据不能公开，谁负责审核申请？需要伦理审批或数据使用协议吗？
5. 是否有代码、脚本或 README 能解释 raw data 到 figure source data 的处理过程？

Common Chinese draft fixes

Recommended bilingual output

When useful, provide English first and Chinese second:

Data Availability
[English statement for submission]

中文核对
- 这句话对应中文含义：[brief Chinese explanation]
- 需要作者确认：[missing accession / repository / ethics condition]

Do not put Chinese explanatory notes inside the final English statement unless the target journal allows bilingual manuscript text.

FAIR Metadata Checklist

Use this file to audit whether a dataset deposit is findable, accessible, interoperable, and reusable enough for a Nature-style submission.

Quick FAIR test

DataCite core fields

Mandatory fields commonly expected for DOI-style dataset records:

• Identifier
• Creator
• Title
• Publisher / repository
• Publication year
• Resource type

Strongly recommended when available:

• contributor and role
• description / abstract
• subject keywords
• funding reference
• related identifiers: manuscript preprint/article, code repository, protocol, previous dataset
• version
• licence / rights
• geolocation or temporal coverage for spatial/temporal data
• language

Dataset README template

# [Dataset title]

## Summary
[One-paragraph description of what the dataset contains and which manuscript results it supports.]

## Files
- [filename]: [contents, format, size, related figure/table]

## Variables and units
[Column/field name] | [definition] | [unit] | [allowed values/missing-value code]

## Methods and provenance
[How data were generated, collected, transformed, filtered, normalised, or aggregated.]

## Software and environment
[Software, package versions, scripts, notebooks, operating system or instrument software when relevant.]

## Access and licence
[Licence, access restrictions, data-use agreement, embargo, or controlled-access process.]

## Citation
[Preferred dataset citation.]

File organization

• Use stable, descriptive filenames instead of local shorthand.
• Keep raw and processed data separate.
• Include a manifest for archives or large multi-file deposits.
• Map source data to exact figure panels and table numbers.
• Preserve units in column names or data dictionaries, not only in manuscript captions.
• Record missing-value codes and filtering decisions.
• Include checksums for large or critical files when the repository does not generate them.

Provenance prompts

Ask the author:

• What instrument, survey, simulation, database, or processing pipeline produced each file?
• Which script or notebook converts raw data into each figure or statistical table?
• Which samples, time points, conditions, or participants were excluded, and why?
• What version of each third-party dataset was used?
• Are there licences, consent forms, data-use agreements, or ethics approvals that limit reuse?
• Has any data been transformed in a way that prevents reconstruction of the raw values?

Licence guidance

• Prefer a standard open licence when data can be public.
• Use the repository's licence field rather than only writing licence text in the manuscript.
• Use CC0 or CC-BY-style terms only when appropriate for the data and institution.
• Do not apply an open licence to third-party or participant data unless the authors hold the right

to do so.

• For code, use a software licence and archive a release when possible.

Final audit

Block submission until these are resolved:

• no Data Availability statement for original research
• no identifier or stable access route for data supporting central conclusions
• sensitive data restriction without access procedure
• third-party data with no source or permission route
• public dataset with no licence or README
• claim that data are in the paper when figure source data are absent
• mismatch between manuscript statement, repository record, and supplementary files

Policy Principles

Use this file when deciding what a Nature-ready data statement must disclose.

Governing rules

• Every original research article needs a Data Availability statement.
• The statement must say what supporting data exist, where they can be found, and any access

conditions.

• The statement must cover data generated by the study and secondary data reused for analysis.
• Public repository deposition is preferred. For community-mandated data types, use the required

repository.

• Reviewers may need access to underlying data and code during evaluation.
• Restrictions are allowed only when they are justified and disclosed. Privacy, consent, endangered

locations, third-party licences, commercial restrictions, and national law are common reasons.

• Restricted data still need a durable access route: named data access committee, institution,

controlled-access repository, application procedure, or responsible group.

• The statement should not hide key evidence in vague language such as "data available upon

reasonable request" unless the reason and process are explicit.

Minimal dataset test

Ask whether an independent reader can inspect or reproduce the paper's central findings from the available material.

Include:

• source data for main figures and key supplementary figures
• raw or sufficiently reusable data, according to community norms
• processed data used for statistics, plots, model training, or validation
• analysis-ready tables if raw data require specialized transformation
• third-party datasets with source, version, date accessed when relevant, and licence/access terms
• representative metadata for restricted datasets, even when records themselves cannot be public

Exclude only when defensible:

• data that were not used to support a result
• purely theoretical work that generated or analysed no dataset
• identifiable human data that cannot be anonymised or shared under consent and law

Availability routes

Use one route per dataset or dataset family.

Data, code, materials, protocols

Data Availability is not a substitute for code, materials, or protocol availability.

• Put custom code in a Code Availability section when the journal separates it.
• Mention code in Data Availability only when it is bundled with the dataset and needed to interpret

files.

• For unique biological materials, reagents, cell lines, plasmids, or model organisms, use

persistent identifiers where available and state distribution restrictions separately.

• For protocols, cite protocol repositories or include enough method detail for reproducibility.

Sensitive and human-participant data

For sensitive data, preserve transparency without breaching consent or law.

State:

• why open sharing is not possible
• whether anonymised, aggregate, synthetic, or representative data can be shared
• where metadata or a summary record is available
• who reviews access requests
• what approval, data-use agreement, or ethics condition applies
• whether access is limited to non-commercial, academic, local-jurisdiction, or qualified users

Avoid:

• naming a single individual as the only durable access route when an institutional route exists
• implying data are available if access depends on impossible or undefined permissions
• promising public release later without a repository, date, and responsible party

Submission-stage checks

Before finalizing, confirm:

• all accession numbers, DOIs, and URLs resolve
• embargoed/private reviewer links work anonymously where required
• restricted data metadata records are public if the records themselves are not
• supplementary files match statement wording
• data citations appear in the reference list where the journal expects them
• no claim depends on unavailable data without explanation

Source notes

• Springer Nature research data policy requires Data Availability statements for original articles

and asks authors to describe available data, location, and access terms.

• Nature Portfolio reporting standards require prompt availability of data, materials, code, and

associated protocols, with restrictions disclosed to editors at submission.

• Scientific Data policy favours repository deposition, especially for primary data, and requires

repository hosting for Data Descriptor datasets.

Repository and Identifiers

Use this file when selecting repositories, checking accession strategy, or writing dataset citations.

Repository decision tree

1. Use a mandated repository when the data type requires it. 2. If no mandate applies, use a discipline-specific, community-recognised repository. 3. If no domain repository fits, use a trusted generalist or institutional repository that provides persistent identifiers and durable metadata. 4. Do not use personal websites, lab websites, ad hoc cloud folders, or unpublished private drives as the only availability route. 5. For very large data, use a repository or institutional infrastructure that can preserve metadata and provide clear access instructions even if bulk files require special transfer.

What a repository record should provide

• persistent identifier: DOI, accession, Handle, ARK, or equivalent stable record
• public landing page with title, creators, abstract/description, repository, date, version, licence
• file list with sizes and formats
• README or data dictionary
• provenance and processing description
• relation to the manuscript and related code
• clear access procedure for restricted data
• versioning or update policy

Common repository categories

Choose according to field norms; this list is not exhaustive.

Always check the target journal and funder because some data types have mandatory repositories.

Identifier rules

• Prefer final public identifiers before submission.
• If the record is private during review, provide an anonymous reviewer link when the repository

supports it.

• Do not cite temporary sharing links as dataset identifiers.
• Include accession numbers exactly as assigned by the repository.
• Use one identifier per coherent dataset record; avoid burying unrelated data under one unclear DOI.
• Version datasets when files change after review or publication.
• If the dataset has a DOI, cite the DOI rather than only the repository URL.

Dataset citation pattern

Dataset references should include the minimum DataCite-style elements:

[Creator(s)] ([Publication year]) [Dataset title]. [Repository]. [Identifier].

Add version when meaningful:

[Creator(s)] ([Year]) [Dataset title], version [version]. [Repository]. [DOI/accession].

For reused public data, cite the dataset in the reference list when the dataset supports conclusions. Mentioning it only in the Data Availability statement may be insufficient.

Repository readiness checklist

Before submission:

• DOI/accession resolves to the intended landing page
• title matches manuscript terminology
• creators and affiliations are correct
• licence is present and compatible with intended reuse
• files open without proprietary software where possible
• README explains columns, units, missing values, transformations, and scripts
• figure source data are clearly mapped to figure panels
• restrictions and access conditions match the manuscript statement
• embargo/private links have been tested outside the author account

Red flags

• "Data available on GitHub" without release DOI or archive
• repository record has no licence
• uploaded zip file has no README or file manifest
• accession exists but is not public, not under embargo, and not available to reviewers
• filenames use local analysis shorthand that readers cannot interpret
• manuscript cites one dataset but results depend on several unlisted secondary sources

Source Basis

Use this file when a user asks why a rule exists, wants primary-source justification, or needs to audit the nature-data skill against real policy sources.

Source map

Official sources

• Springer Nature, Research data policy:

<https://www.springernature.com/gp/journal-policies/15369670>

• Springer Nature, Data availability statements:

<https://www.springernature.com/gp/authors/research-data-policy/data-availability-statements>

• Springer Nature, Data repository guidance:

<https://www.springernature.com/gp/authors/research-data-policy/recommended-repositories>

• Springer Nature, Sensitive data:

<https://www.springernature.com/gp/authors/research-data-policy/sensitive-data>

• Nature Portfolio, Reporting standards and availability of data, materials, code and protocols:

<https://www.nature.com/nature-portfolio/editorial-policies/reporting-standards>

• Example Nature Portfolio journal reporting standards page:

<https://www.nature.com/npj2dmaterials/editorial-policies/reporting-standards>

• Nature Research, Data availability statements and data citations policy FAQ:

<https://www.nature.com/documents/nr-data-availability-statements-data-citations-faqs.pdf>

• Scientific Data, Data policies:

<https://www.nature.com/sdata/policies/data-policies>

• Wilkinson et al. 2016, The FAIR Guiding Principles for scientific data management and stewardship:

<https://www.nature.com/articles/sdata201618>

• DataCite Metadata Schema:

<https://schema.datacite.org/>

Notes for future updates

• Check target journal instructions first because Nature Portfolio journals can add field-specific

requirements.

• Check DataCite's latest schema before naming version-specific fields. As of 2026-05-01, the

DataCite schema landing page lists Metadata Schema 4.7 as the latest release.

• Keep this file as a source map, not a long policy mirror. Link to official pages rather than

copying full policy text.

Statement Patterns

Use these patterns as starting points. Replace bracketed fields with verified information. Delete any sentence that does not apply.

For Chinese users, treat the Chinese line under each pattern as author-facing guidance, not as submission text. Submit the English statement unless the journal explicitly asks otherwise.

Public repository, single dataset

The [raw/processed/source] data supporting the findings of this study are available in
[Repository] under accession [ACCESSION] / at [DOI or persistent URL]. The deposited record
contains [brief contents: e.g. raw measurements, processed tables, figure source data, metadata
and analysis inputs].

中文对应：本研究的原始/处理后/源数据已存储在某个正式仓库，并有登录号、DOI 或永久链接。

Public repository, multiple datasets

The datasets generated in this study are available as follows: [dataset family 1] in
[Repository] under [DOI/accession]; [dataset family 2] in [Repository] under [DOI/accession];
and figure source data in [Repository/Supplementary Data file] under [identifier or file name].

中文对应：不同类型数据分别放在不同仓库或文件中，需要逐一说明，不能笼统写“数据见附件”。

Data in paper and supplementary files only

Use only when the supporting dataset is genuinely small and fully represented in the article, source data, or supplementary files.

All data supporting the findings of this study are included in the paper, its Supplementary
Information, and Source Data files. [Name exact Supplementary Tables/Data files when possible.]

中文对应：只有当支撑结论的数据确实都在正文、补充材料和 Source Data 中时才这样写。

Reused public data

This study used publicly available [dataset name/type] from [Repository or source], available under
[DOI/accession/stable URL]. We used [version/release/date accessed, if relevant]. No new primary
[data type] data were generated for this part of the analysis.

中文对应：使用公开数据库时，需要写清数据库名、版本/发布日期/访问日期和编号，并引用数据集。

Mixed generated and reused data

Data generated in this study are available in [Repository] under [DOI/accession]. Public datasets
reused in the analysis were obtained from [source 1, identifier/version] and [source 2,
identifier/version]. Source data for [figures/tables] are provided in [location].

中文对应：自己产生的数据和复用的公开数据要分开写，避免让读者误以为所有数据都是本研究产生。

Controlled-access human or sensitive data

The [data type] data supporting this study are not publicly available because [privacy, consent,
legal, ethical or security reason]. A metadata record is available at [repository/accession, if
available]. Qualified researchers may request access from [data access committee/institutional
office/repository procedure] at [contact or URL]. Access requires [ethics approval/data-use
agreement/other conditions] and will be reviewed according to [policy or committee name].

中文对应：涉及人类参与者、隐私或伦理限制时，不能只写“因隐私不可公开”；还要写申请路径和审核条件。

Third-party or licensed data

The [data type/name] data used in this study were obtained from [third-party provider] under
licence and are not publicly redistributable by the authors. Requests for access should be directed
to [provider/contact/URL]. Derived data that can be shared are available in [repository] under
[DOI/accession], subject to [licence or restriction].

中文对应：第三方授权数据不能由作者重新分发时，要说明数据所有者和读者应向谁申请。

Commercially restricted data

The [data type] data are subject to commercial restrictions and cannot be made publicly available.
Requests for access may be directed to [company/data owner/contact or URL] and are subject to
[approval/licence/payment/confidentiality terms]. The authors provide [summary statistics,
metadata, synthetic data, or source data] in [location] to support interpretation of the results.

中文对应：企业或商业数据不可公开时，需要说明商业限制、申请对象，以及是否有汇总数据或元数据可公开。

Embargoed data

Use only when the repository supports embargo and the journal permits it.

The [data type] data have been deposited in [Repository] under [DOI/accession] and are under
embargo until [date/event]. Reviewers can access the data using [private reviewer link or
repository access route]. The data will become publicly available at [DOI/accession] when the
embargo ends.

中文对应：如果数据暂时不公开，必须已有仓库记录、审稿访问方式和明确解封时间或条件。

Request-based access with justified restriction

The [data type] data are not publicly available because [specific reason]. Requests for access may
be sent to [institutional group/contact route], and will be considered for [eligible purpose/users]
subject to [approval, agreement, or legal condition]. [Public metadata/aggregate data/source data]
are available at [location].

中文对应：“合理请求”只有在说明原因、接收机构、审核条件和可公开元数据后才可接受。

No datasets generated or analysed

Use sparingly.

No datasets were generated or analysed during the current study.

中文对应：只有确实没有生成或分析任何数据时才能使用，经验研究通常不适用。

For theory papers, be more specific:

This work is theoretical and does not generate or analyse empirical datasets.

Anti-patterns to revise

Audit questions

• Which result would fail if this dataset were unavailable?
• Is the route durable beyond the corresponding author's current email address?
• Can a reader tell what each identifier contains?
• Are restrictions specific enough for an editor to judge them?
• Are reused datasets cited, not merely mentioned?

Chinese-user operating mode

When the user writes in Chinese, provides a Chinese manuscript note, or asks for "中文对应", "中英对照", "数据可用性声明", "数据获取声明", "原始数据", "数据存储库", or "受限数据":

• Accept Chinese input naturally, but draft the final submission-ready statement in English unless the user explicitly asks for Chinese only.
• Preserve a short Chinese explanation of unresolved decisions when it helps the author act.
• Translate intent, not wording. Chinese phrases such as "可向通讯作者索取" are usually too vague for Nature-style English unless the restriction and access process are specified.
• Convert Chinese repository/status descriptions into precise publication terms:

数据可用性声明 -> Data Availability; 原始数据 -> raw data; 处理后数据 -> processed data; 源数据 -> source data; 补充材料 -> Supplementary Information; 受限数据 -> restricted data; 合理请求 -> reasonable request, only with reason and review route.

• Use references/chinese-author-alignment.md for Chinese terminology, common CN-to-EN failure modes, and bilingual intake questions.

Default stance and source hierarchy

Use this skill to turn a manuscript's supporting data into a transparent, Nature-ready data availability package: statement text, repository plan, dataset citations, and missing-information flags.

The governing policy layer is Springer Nature / Nature Portfolio data policy. The implementation layer is FAIR data practice and DataCite-style citation metadata.

Default stance

• Treat the Data Availability statement as a link between the paper's claims and the evidence needed to inspect, reproduce, or reuse them.
• Do not invent DOIs, accession numbers, repository names, licences, embargo dates, ethics approvals, access committees, or data-use conditions.
• Prefer public, discipline-specific repositories. Use generalist or institutional repositories only when no suitable community repository exists.
• Describe both newly generated data and reused third-party data.
• If data cannot be openly shared, state why, who controls access, how requests are evaluated, and what metadata or representative data can still be public.
• Separate data, code, materials, and protocols unless the journal asks for a combined availability section.
• Keep this skill focused on availability and metadata. Do not rewrite methods, analyze statistics, or polish the manuscript unless the user asks for those tasks separately.
• Flag "available upon request" as weak unless there is a specific legal, ethical, commercial, or third-party restriction.

Source hierarchy

Use sources in this order:

1. Target journal instructions and submission system requirements. 2. Nature Portfolio / Springer Nature data, code, materials, and reporting policies. 3. Repository-specific requirements and domain community standards. 4. FAIR principles and DataCite metadata practice.

If a policy detail may have changed, verify the current journal page before giving final submission advice.

Workflow and output format

Workflow

1. Identify the target journal and article type. If journal-specific instructions conflict with this skill, follow the journal. 2. Inventory every dataset needed to support the main and supplementary results: generated raw data, processed data, figure source data, secondary data, software outputs, models, tables, images, and files underlying statistical analysis. 3. Classify each dataset into one access route: public repository, controlled access repository, within paper or supplement, reused public source, third-party restricted, available on justified request, or not applicable. 4. Choose repository and identifier strategy before drafting text. Prefer DOI, accession number, Handle, ARK, or stable repository record over personal websites and temporary cloud links. 5. Draft the Data Availability statement using explicit dataset-to-location mapping. 6. Add formal dataset citations for public data that support conclusions. 7. Run the FAIR and metadata audit before finalizing. 8. Return ready-to-paste statement text plus any unresolved fields the author must confirm.

Output format

Unless the user asks for another format, return:

Data Availability
[ready-to-paste statement]

Repository and citation actions
- [specific actions or "None"]

Missing information / risk flags
- [specific flags or "None"]

中文核对
- [用中文列出作者需要确认的字段或 "无"]

When auditing an existing statement, lead with blocking issues first, then provide a revised version.