Nature Skills is a collection of reusable skills that help AI agents handle academic writing and scientific visualization. Researchers and AI-assisted scholars use it to turn research tasks into repeatable workflows and usable outputs. The catalogue entries are skills from this collection.
Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add Yuan1z0825/nature-skills --skill nature-ref-verifiergit clone --depth 1 https://github.com/Yuan1z0825/nature-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/yuan1z0825/nature-skills/nature-ref-verifier)<a href="https://agentmods.dev/skills/yuan1z0825/nature-skills/nature-ref-verifier"><img src="https://agentmods.dev/badge/skills/yuan1z0825/nature-skills/nature-ref-verifier/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/yuan1z0825/nature-skills/nature-ref-verifier"><img src="https://agentmods.dev/badge/skills/yuan1z0825/nature-skills/nature-ref-verifier.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- Socket pass
- Snyk warn
- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Data Exfiltration · line 45 Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.Fix: Verify the destination URL is trusted and necessary. Remove or replace with documented APIs. Ensure no secrets, tokens, or PII are transmitted.
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00096 | $0.02834 |
| Opus 5 | $0.00048 | $0.01417 |
| Sonnet 5 | $0.00019 | $0.00567 |
| Haiku 4.5 | $0.00010 | $0.00283 |
Grade A, and why
nature-ref-verifier scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 13d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 197 lines — stays where its author put it; the contents beside it link to each section on GitHub.
nature-ref-verifier — 学术参考文献多源验证技能
需要处理批量差异、反复出现的字段错误、DOI/标题冲突、作者顺序异常、页码漂移或
报告措辞时，读取 references/common-patterns.md。其他任务不要预加载该文件；
manifest.yaml 记录了同一按需路由，新增资源时必须同时在本文件中写明加载条件。
触发词
verify references、校验文献、check references、核对参考文献、文献验证、ref check
适用场景
- 开题报告/论文提交前，对全部参考文献做最终核查
- 收到审稿意见指出引用错误时，精准定位问题
- 从旧项目迁移参考文献到新论文时，批量验证有效性
- Zotero 库定期健康检查
工作流
Step 1: 解析输入
支持三种输入格式：
A. 整篇论文/开题报告
- 提取所有
[N]编号的参考文献列表 - 自动识别每条的 DOI（如有）
B. 单条参考文献
- 用户直接粘贴一条引用文本
- 或传入 Zotero item key
C. BibTeX 文件
- 直接解析
.bib，逐条校验
Step 2: 多源并行查询
Step 2.0（最便宜、最先做）：DOI 可解析性检查。 对每条有 DOI 的文献先请求 https://api.crossref.org/works/<DOI>：404 → DOI 本身错误（高度疑似错号）；200 → 比对返回记录的标题/作者/卷期页与条目，不一致即"DOI 张冠李戴"。单次请求即可发现实践中最常见的一类错误（一次 35 条核查中 7/9 的错误属于此类）。
对每条文献，并行查询以下来源（可用哪些取决于环境配置）：
| 来源 | 适用文献 | 典型覆盖率 |
|---|---|---|
| Crossref | 有 DOI 的外文期刊 | ~70%（IEEE 老文献、中文期刊常有缺失） |
| IEEE Xplore | IEEE 期刊/会议 | IEEE DOI 404 时的备选 |
| WebSearch (Bing/Google) | 无 DOI 或来源缺失的文献 | 兜底方案 |
| CNKI/万方 (kimi-datasource / webbridge) | 中文期刊、学位论文 | 中文文献必查 |
| Zotero 本地库 (zotero-mcp) | 已入库文献 | 快速比对已有数据 |
查询策略： 优先用 DOI 查 Crossref，失败后降级用标题+作者搜 Web。
批量策略： 超过 20 条时按 10–15 条一组拆给并行子代理，每组独立查询后汇总——35 条规模实测 3 个并行子代理即可在一次往返内完成，串行会慢一个数量级。
Step 3: 字段级对比
对每条文献执行以下对比矩阵，按严重程度分为三级：
🔴 Critical（必须修正）
| 检查项 | 说明 | 典型案例 |
|---|---|---|
| 作者姓名/顺序不一致 | 第一作者不同、顺序颠倒、完全编造 | Smogavec→Leckebusch、Vainikainen→Mikhnev |
| 作者漏人 | 条目作者数少于出版方记录（常见于 5+ 作者文献只录前 3-4 人） | TGRS 2019 PCA 文漏第 5 作者 Ge；TAES 2021 UAV 文漏第 5 作者 Kim |
| 页码差异 ≥5 | 页码完全不匹配 | Noon: 1309-1319→1444-1449 |
| 文章号字母误判 | 文章号含字母被写成形近数字（T/7、l/1、O/0） | SPIE 101880T 误作 1018807 |
| 标题核心词不一致 | 非大小写/标点差异 | Iwasaki T→Wu K H、Zhang W→Zhu M H |
| DOI 指向不同论文 | DOI 存在但标题/作者不对应 | Abidi 1995: 10.1109/4.482157→10.1109/4.482187 |
| DOI 张冠李戴 | DOI 可解析但解析到另一篇论文（比对标题作者即可发现） | JAG 2017 SVD 文 DOI 指向同卷 ERT 论文；AST 2022 旋翼文 DOI 指向涡轮噪声论文 |
🟡 Warning（建议核对）
| 检查项 | 说明 | 典型案例 |
|---|---|---|
| 卷年 ≠ DOI 年 | 卷归属年份与 DOI 编号年不一致 | Dadrass 卷2025/上线2024、Ni 卷2022/DOI含2020 |
| Early Access 元数据漂移 | IEEE 等期刊 Early Access 转正式出版后卷/期/页/年会变化，以 CrossRef 现值为准 | TGRS autoencoder 文 59(12):10437-10449,2021 → 60:1-13,2022 |
| 作者中间名缺失/多余 | "Rabiner L" vs "Rabiner L R" | 一般不影响检索 |
| 期号缺失 | 脚本有卷无期 | Smogavec 2025: 17→17(10) |
| 页码偏差 ≤4 | 细微差异 | Leckebusch 15-24→15-25 |
| 出版社名写法差异 | "IET" vs "IEE" | 历史名称变更 |
What ships with it
5 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 13d ago First seen · 197 lines · 96 tokens per session scan A 6a975074d7a1
nature-ref-verifier is a skill published in the GitHub repository Yuan1z0825/nature-skills (40,913 stars, last pushed 2d ago), licensed Apache-2.0. It adds 96 tokens to every session and 2,834 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
literature-searcher
Search CrossRef, OpenAlex, PubMed, Semantic Scholar, and optional Scopus; deduplicate results, download open-access PDFs by DOI, classify papers, monitor new results, and analyze coverage. Use when asked to search literature, monitor a topic, download an open-access paper, classify papers, or analyze literature gaps.
academic-research
Search academic papers, scholarly articles, and research publications through SandBase. Use when asked for literature review, academic citations, scholarly research, paper discovery, or scientific evidence gathering.
academic-trend-research
Research emerging academic trends by combining scholarly databases, web sources, and news to identify rising research areas, breakthrough papers, and shifting scientific consensus. Ideal for R&D teams tracking frontier science.
using-cesiumjs-skills
Use when starting any conversation involving CesiumJS development - provides orientation on available domain skills and how they activate.
latex-empirical-tables
Set up, format, fix, and clean up LaTeX regression and estimation tables in empirical economics or finance papers. Use when (a) creating a .tex, results.tex, main.tex, or preamble to display regression or estimation output from Python (pyfixest), Stata, or R — including any mention of estout, esttab, estauto, estwide…
cite-placement
Place pre-screened literature citations into a LaTeX or Word manuscript, or restyle the citations already in one. Three modes: (1) inline placement — inline \cite{}/\citet{}/\citep{} with a compiled references.bib, for author-date journals (APA, MLA, Harvard, Chicago author-date, IEEE, Vancouver); (2) footnote…