What does this skill do, and when should you use it?
This skill helps researchers prepare, audit, or revise data availability sections for Nature-family manuscripts, including repository selection, dataset citations, and FAIR metadata audits. It uses a layered static/dynamic structure and an eight-step workflow to inventory datasets, classify access routes, and draft ready-to-paste statements while respecting journal-specific guidelines. The skill also supports Chinese-speaking authors by accepting Chinese input and converting it into precise English academic wording.
The skill loads core guidance files and runs an eight-step workflow: identify the journal, inventory every supporting dataset, classify each into one of seven access routes (public repository, controlled access, within paper, reused public, third-party restricted, justified request, or not applicable), choose repository and identifier strategy, draft the statement with explicit dataset-to-location mapping, add formal dataset citations, perform a FAIR/metadata audit, and return ready-to-paste text plus unresolved fields. It does not invent DOIs, accession numbers, repository names, licences, embargo dates, ethics approvals, access committees, or data-use conditions; it flags 'available upon request' as weak unless a specific restriction exists.
- A Chinese-speaking researcher preparing a Nature manuscript needs accurate bilingual data availability wording and a structured workflow.
- An author wants to audit an existing data availability statement before submission to ensure every dataset has a clear access path and citation.
- A researcher needs help selecting the right repository (public vs. controlled access) and obtaining proper accession numbers for datasets.
- An author wants to generate a FAIR metadata checklist to comply with open data mandates.
- A student writing a general academic paper needs a data availability statement template, not specific to Nature.
How do you install this skill?
- Static review only, not executed; reliability, effectiveness, and verifiability scores are based on file inspection and uncertain.
- The skill depends on static/ and references/ files not included in this review, which may affect full assessment.
- Publisher identity is unverified; users should consider source credibility.
- The skill lacks explicit user confirmation steps; users should ensure they are informed and confirm critical information.
- Local filesystem
This skill is part of the nature-skills monorepo. Install via npx skills: npx skills add Yuan1z0825/nature-skills --agent codex --skill nature-data --yes --copy. For manual installation, keep the full skill directory (SKILL.md, manifest.yaml, static/, references/) — do not copy only SKILL.md.
tmp="$(mktemp -d)"
git clone --depth 1 https://github.com/Yuan1z0825/nature-skills.git "$tmp"
mkdir -p ~/.claude/skills
cp -R "$tmp/skills/nature-data" ~/.claude/skills/
rm -rf "$tmp"Generated from the source repository and skill path; it copies only this skill's folder. If the author's install steps above differ, follow those first. To scope it to one project, replace ~/.claude/skills with that project's .claude/skills.
How do you use this skill?
Once installed, send your agent any of these to trigger it:
- Please write a Data Availability statement for my Nature manuscript
- Generate a FAIR metadata checklist for my datasets.
After installation, describe your task in a client that supports Agent Skills (e.g., Claude Code or Codex), such as: 'Please write a Data Availability statement for my Nature manuscript' or 'Generate a FAIR metadata checklist for my datasets.' The skill automatically loads core files, runs the workflow, and returns a ready-to-paste statement.
What are this skill's strengths and limitations?
- Tailored to Nature-family journals, meeting their academic writing and data-sharing expectations.
- Layered structure keeps the core lean and loads detailed references only when needed, reducing overhead.
- Supports Chinese-speaking authors by converting Chinese requests into precise English statements and adding a Chinese validation block.
- Explicitly forbids inventing DOIs, accession numbers, repository names, licences, embargo dates, or ethics approvals, reducing fabrication risk.
- Includes a FAIR metadata audit to help authors meet data-sharing compliance.
- Marked as Draft status — rules are defined but not tested on real cases, so edge cases may exist.
- Requires complete paper and data information; otherwise, output will contain unresolved fields.
- Depends on reference files (e.g., policy-principles.md, repository-and-identifiers.md) that must be kept up to date.
- Installation requires keeping the full directory structure, which may be non-obvious for some users.
- Provides text output only; does not generate charts or code.
How does this skill compare with similar options?
Side by side with related skills; every score comes from the same FSRS standard.
| Skill | FS score | Stars | Last updated | License |
|---|---|---|---|---|
| Nature Data Availability Skill this page | 53 · Use with care | ★ 47k | 3d ago | Apache-2.0 |
| Nature/CNS Citation Skill | 56 · Use with care | ★ 47k | 3d ago | Apache-2.0 |
| Shuorenhua: Chinese AI-Tone Cleanup Skill | 65 · Recommended | ★ 2k | 11d ago | MIT |
| Human Writing | 64 · Recommended | ★ 4k | 2mo ago | MIT |
| Academic Humanizer | 55 · Use with care | ★ 1.9k | 3mo ago | NOASSERTION |
How did FollowSkills review this skill?
Evidence: the skill emphasizes not inventing DOIs, accessions, repositories, or access conditions and requires user confirmation for missing information, reflecting transparency and avoiding overclaiming. No excessive permissions or dangerous operations; dependencies are local reads and text generation, no network calls or external side effects. However, explicit user confirmation steps (only mentioned in boundaries) and rollback mechanisms are lacking, and publisher identity is unverified. Deduction: incomplete permission/confirmation, missing rollback.
Evidence: the skill structure is clear with a router SKILL.md loading static and references, and manifest.yaml defines loading rules. But no tests or execution evidence are present, and whether the key path (workflow) reproduces reliably is unknown. Handling of abnormal inputs is not specified, failure feedback quality unknown. Due to static review and lack of test evidence, deduct: no tests, thin edge-case handling.
Evidence: use case is clear (prepare, audit, or revise Data Availability statements), triggers are precise including Chinese trigger words (数据可用性声明, etc.). Environment fit is good: supports Chinese users, bilingual, no external network dependency. Boundaries clear: does not invent data, provides checklist when missing. However, concrete input/output format examples are missing, trigger precision unverified. Deduct: limited boundary evidence, trigger precision unverified.
Evidence: clear information architecture with README.md and README_EN.md, static/dynamic layering, version 2.0.1 in manifest.yaml, Apache-2.0 license. But changelog, known-limitations disclosure, and explicit maintenance responsibility are missing; update path not clear. No mention of update mechanism; publisher identity unknown. Deduct: unclear version history and update path.
Evidence: skill has clear goals and output format (paste-ready English statement, mapping table, FAIR checklist), but no actual run results or examples verifying output quality. Static review cannot confirm ability to complete core task; evidence of marginal value lacking. Due to no execution or verification, deduct: output direct usability lacks evidence.
Evidence: provides reference file (references/source-basis.md) but not included in this review, no test suite or CI for this skill. No third-party execution evidence, cannot verify core workflow reliability. Deduct: no independent verification, limited source traceability.
Open a dimension to read why it scored that way
Evidence confidence：Low — Mostly static review, author material or a limited demo; useful for discovery, not high-risk decisions.
See the full review method →