Scientific Review

Conduct systematic literature reviews on PubMed with enforced full-text verification and independent audit loops. Use when the user asks for a comprehensive literature search, systematic review, evidence synthesis, or deep research on a biomedical topic. Trigger when the user mentions "literature review", "systematic review", "调研", "文献综述", "汇总分析", "comprehensive search", "what does the literature say", or asks to survey published research on a gene, pathway, disease mechanism, or clinical question. Make sure to use this skill whenever the user wants thorough evidence synthesis — do not skip it

7 starsResearch & Science

Install

skills CLI (Claude Code, Cursor, Codex and more)

npx skills add wuliRothschild/scientific-review --skill scientific-review

Manual install

git clone --depth 1 https://github.com/wuliRothschild/scientific-review.git /tmp/scientific-review
mkdir -p ~/.claude/skills
cp -r /tmp/scientific-review/. ~/.claude/skills/scientific-review

Personal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.

From the SKILL.md

你是一位生物医学文献研究助手。按以下 Phase 执行——每步有明确输出。

PubMed 摘要 ~250 词，必然遗漏关键信息。摘要检索是**起始层**而非全部方法——关键结论必须由全文层证据支撑，"摘要未列出"不等于"论文不涉及"。

***

必须严格串行执行完当前步骤后，方可汇总信息整理，再决策如何执行下一步骤的操作，禁止为了效率和省时并行或跳过当前Phase或操作，无论耗时多久。若存在任何跳过当前阶段或phase或操作执行后续阶段或phase或操作的现象，则所有交付和结论视作无效，整个任务视作失败，所有基于报告得出的结论或解释均为学术造假和不端。用户将在任务结束后检查是否存在任何跳过当前阶段或phase或操作执行后续阶段或phase或操作的现象或痕迹。

## Phase 1 — 问题分类与维度规划

### 1a. 问题类型

| 类型          | 核心需求     | 检索重点        | 报告侧重           |
| ----------- | -------- | ----------- | -------------- |
| **机制/通路**   | X 如何调控 Y | 机制研究、通路分析   | 因果链、直接 vs 间接证据 |
| **临床疗效**    | 治疗/诊断表现  | RCT、荟萃分析、队列 | 效应量、NNT、置信区间   |
| **方法比较**    | 哪种更好     | 头对头比较、设计论文  | 定量对比表、强度层级     |
| **图谱/资源**   | 有什么数据/工具 | 资源论文、联盟出版物  | 覆盖度、访问方式、局限    |
| **标记/稀有细胞** | 细胞身份鉴定   | 标记查询、聚类注释   | 标记面板、谱系、功能证据   |
| **流行病/趋势**  | 患病率、风险因素 | 大型队列、登记研究   | 率、趋势、人群特征      |

### 1b. 信息维度分析

明确回答此问题需要哪几类不同证据。常见维度对：

* 应用证据 + 方法学证据

* 人类数据 + 动物/体外数据

* 原始研究 + 系统综述/指南

### 1c. 输出声明

```
问题类型：[机制/临床/方法/图谱/标记/技术]
信息维度：[维度1, 维度2, ...]
检索难度：[直接/全文敏感/术语依赖]
```

检索难度说明：

* **全文敏感**：大多数生物医学问题的默认值。定量数据、图表、方法细节、补充材料在 PubMed 摘要中不出现，自动升级到全文检索。仅当问题明确为"是否有 X 证据"的快速确认型时，才视为直接型。

* **术语依赖**：标准名与用户用词不同——Phase 2d 术语发现循环不可跳过。

* 全文敏感型 → Phase 2 查询构建时，优先使用 PMC 可及性筛选，扩大命中全文的概率。

* 全文敏感型 + 术语依赖型 → Phase 3 全文获取为强制路径，不入降级逻辑。

***

## Phase 2 — 查询构建、执行与首轮审计

### 2a. 术语矩阵

基于领域知识，为问题涉及的每个概念列出标准名、同义词/别名、MeSH 词与用户用词：

| 概念 | 标准名 | 同义词/别名 | MeSH 词 | 用户用词 |
| -- | --- | ------ | ------ | ---- |

### 2b. 构建并执行

Read the full skill on GitHub

Details

Skill name: scientific-review
Repository: wuliRothschild/scientific-review
License: Apache-2.0
Repo updated: 29 October 2026

FAQ

What does the Scientific Review skill do?

How do I install the Scientific Review skill?

skills CLI (Claude Code, Cursor, Codex and more): npx skills add wuliRothschild/scientific-review --skill scientific-review. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.

Is Scientific Review popular?

It lives in wuliRothschild/scientific-review, which has 7 GitHub stars and 0 forks.

Related skills

Beautsgo Booking

BeautsGO/beautsgo-booking

Book appointments at 1300+ top-rated Korean dermatology & plastic surgery clinics in Seoul, Busan, Jeju directly from your AI assistant. Supports laser, injection, Botox, skin boosters, double eyelid, rhinoplasty, anti-aging, acne treatment in Chinese/English/Japanese/Thai. Keywords: Korea medical tourism, Seoul skin clinic, Korean dermatology, plastic surgery Korea, 韩国医美预约, 韩国皮肤科, 韩国整形外科, 首尔美容院, 医疗旅游韩国, 韩国整容, 水光针预约, 肉毒素预约, 韩式双眼皮. 热门医院: 江南 — 梅宗德/Barog/JD皮肤科/hev赫熙/鹿美人/secret希瑞特/金泰拉/伊美芝/爱妮/美LAB/Oganacell奥嘉娜/reberry/Shinebom/ELEV/Pind/GD医院/ID医院/陶瓷医院; 明洞 — UMI优美/reberry/丹雅/lijin/可丽/daybeau/奥缇娜/本思;

Research & Science

Pedagogical Grounding Engine

edu-ai-builders/pedagogical-grounding-engine

将一篇或多篇学术论文，通过"教学法约束推理"转化为有理论根基的产品设计产物。核心是构建"学术机制 → 约束交互 → 可解释推理链"的中间层，再导出多种格式的设计产物。本skill是一个 Pedagogical Grounding Engine（教学法落地引擎）：它不是论文摘要工具，而是把教学法意图（pedagogical intent）编译为可执行的系统设计结构。PRD、System Prompt、Workflow 都是下游导出格式，核心产物是 Traceability Map——每个设计决策都有明确的教学依据、认知机制和可追溯的设计理由。当用户有以下意图时触发本skill： - "把这篇/这几篇论文做成产品/app/提示词/工作流" - "paper to PRD / paper to prompt / paper to product" - "pedagogical grounding / 教学法落地" - "我想基于学习科学设计这个功能/系统提示词" - "从论文提取知识图谱/三元组/TPACK分析" - "我有几篇论文，帮我提炼成一个设计方案" - "帮我把教学法嵌入到产品/提示词中" - "生成 traceability 图" / "把这篇论文可视化" - "我有一个教学场景/教育想法，帮我设计成系统"（不一定有论文）支持 Commands 模式：流水线

Research & Science

Collage Broll Minimax

jiangge0804-crypto/collage-broll-minimax

将约 5 秒口播文稿、观点句或抽象概念做成高级 editorial halftone paper-collage / 半调纸拼贴 B-roll。用户说“拼贴动画”“collage b-roll”“纸拼贴 b-roll”“半调拼贴”“拼贴风格配画面”“用这段文稿做拼贴动画”“gbro-collage-broll”“collage-broll-minimax”，或希望把一句文稿转成拼贴视觉隐喻时，必须使用此 skill。强制采用三阶段审批：先只提视觉隐喻，用户确认后才生成彩色拼贴静帧，静帧再次确认后才默认调用 MiniMax H3（海螺 3.0，platform.minimaxi.com 国内直连）生成首尾帧组装动画。默认视频模型固定为 MiniMax-H3；只有用户明确指定其他模型时才切换。

Research & Science

Redacta

nickjlamb/redacta

Pseudonymises medical and clinical documents by replacing patient identifiers with labelled tokens (e.g. [PATIENT_NAME_1], [NHS_NUMBER_1], [DATE_OF_BIRTH_1]) so the text can be safely processed by AI or shared, with clinical meaning intact. Combines a deterministic pattern layer (NHS numbers with Modulus-11 validation, UK National Insurance numbers, dates of birth, UK postcodes, phone numbers, emails, hospital/MRN numbers) with contextual reasoning for patient names, postal addresses and identifying ages, then returns the redacted document plus a redaction report. Use when the user wants to re

Research & Science

Alternative Coffee Intelligence

ElmatadorZ/alternative-coffee-claudeskill

Alternative Slowbar coffee intelligence — scientific analysis from seed to cup using First Principle + System Thinking. Covers: varieties, terroir, cultivation, disease diagnosis, processing, roasting physics/chemistry, extraction science, brewing, business strategy, storytelling. ALWAYS trigger for coffee-related questions requiring scientific depth. DO NOT trigger for simple recipes or general food questions. Style: science-driven, farmer-first, quality-obsessed.

Research & Science

Report

MarcinDudekDev/claude-report-skill

Generate self-contained HTML reports instead of long terminal output. For task completions, research, audits, comparisons, and blockers.

Research & Science

Browse the most popular Claude skills, all research & science skills, or write your own with the skills generator.