# Add Nature/CNS Citations to Your Manuscript

Academic writers struggle to find high-impact Nature, Science, and Cell citations to support claims in their manuscripts. This skill splits manuscript text into segments and searches only accepted flagship and subjournal titles to produce reference-manager-ready citation files.

## Install

```bash
npx skillstore add yuan1z skill, refactored into static/dynamic layers/yuan1z0825-nature-citation
```

## Metadata

- - Slug: yuan1z0825-nature-citation
- - Version: 2.0.0
- - Author: Yuan1z skill, refactored into static/dynamic layers
- - GitHub username: Yuan1z0825
- - License: MIT
- - Repository: https://github.com/Yuan1z0825/nature-skills/tree/main/skills/nature-citation
- - Ref: main
- - Supported tools: Claude, Codex, Claude Code
- - Risk level: low
- - Risk factors: external\_commands, network, filesystem, scripts
- - Quality score: 84
- - Quality tier: gold
- - Public page: https://skillstore.pages.dev/skills/yuan1z0825-nature-citation
- - Manifest: https://skillstore.pages.dev/api/skills/yuan1z0825-nature-citation/manifest

## Capabilities

- Splits manuscript paragraphs into citable segments using configurable character limits
- Searches Crossref metadata API restricted to Nature Portfolio, AAAS Science family, and Cell Press journals
- Filters candidate papers by publication year range \(--from-year, --to-year\)
- Exports results in EndNote \(.enw\), RIS, and Zotero RDF formats
- Generates an HTML browser for interactive review and selection of candidate citations
- Supports Chinese-language academic writing workflows with bilingual segment handling

## Use Cases

- Adding citations to a journal manuscript draft: A graduate student pastes their manuscript text and receives a ready-to-import EndNote file with verified Nature/CNS candidate references for each claim.
- Building a reference list for a review article: A researcher preparing a review uses the skill to systematically find supporting citations from top-tier journals across multiple paragraphs of their draft.
- Chinese-language academic writing with CNS references: A Chinese-speaking researcher writing in Chinese gets citations with bilingual segment handling while searching the same Nature/CNS journal scope.

## Prompt Templates

### Basic citation request

```
Add Nature and Science citations to the following paragraph: [paste your text here]
```

### Scoped search with year filter

```
Find CNS citations for this manuscript section, restricted to papers published between 2020 and 2025: [paste text]
```

### Chinese-language manuscript

```
帮我为以下论文段落添加Nature系列和CNS子刊的引用：[粘贴中文段落]
```

### Batch processing long article

```
Process this 5000-word manuscript in batches and export all unique CNS citations as a single RIS file: [paste or reference file]
```

## Limitations

- All candidates are tagged 'metadata-only' and must be verified against abstracts before citing
- Does not retrieve full-text content or verify claims against source material
- Citation matching relies on Crossref metadata which may not always reflect the most recent publications
- Limited to the predefined journal scope \(Nature/CNS/Cell families\); cannot search arbitrary journals

## Best Practices

- Always verify candidate citations against the abstract or publisher page before including them in your manuscript
- Use the --from-year and --to-year flags to restrict searches to relevant publication periods for your field
- Start with a smaller scope \(--scope flagship\) before expanding to subjournals to reduce false positives
- For long manuscripts, use --batch-size to process segments in chunks and avoid rate limiting from Crossref

## Anti Patterns

- Do not cite a paper solely because its title is related to your claim without reading the abstract
- Do not rely on this skill for journals outside the Nature/Science/Cell families - it will return no results
- Do not use the output as a final reference list without manually verifying each citation's relevance and accuracy

## Security Audit

- - Safe to publish: true
- - Audited at: 2026-06-24T05:55:57.25\+00:00
- - Summary: Evaluated 308 static findings. The vast majority are false positives: 'Ruby/shell backtick execution' flags are Markdown code formatting \(not actual execution\), 'weak crypto' flags match 'md5' in documentation contexts, 'certificate/key' flags refer to citation keys, and 'system reconnaissance' flags are CSS class names. The script makes legitimate HTTP requests to api.crossref.org for citation metadata search and writes output files locally. One low-severity XSS risk exists in the self-contained HTML browser output where template literals render Crossref API data via innerHTML.

## Stats

- - Views: 0
- - Downloads: 2
- - Favorites: 0
- - Popularity score: 0