Light — 全流程科研技能包:从文献调研、数据处理、创新提案、方案设计、实验分析,到论文写作润色、图表绘制、排版投稿、审稿返修、软著专利、PPT 与竞赛材料,配套庞大可核查的科研知识库与对抗式自检流程。
后端代码编写、逻辑强、安全性高、可读性好、版本控制、代码审查。当任务需要写实验代码、模型代码、数据处理代码、可视化代码、后端接口或系统逻辑时使用。要求逻辑清晰、安全、可读、可维护、便于复现/扩展/部署。支持 Git 版本管理、代码审查、注释规范、README、依赖管理、环境配置、运行说明与项目结构整理。
论文引用规划、审查与多格式生成。当用户需要处理参考文献、引用、bibtex 时使用。审查引用的关联度、真实性、权威性、时效性、数量、中外占比,是否引用了经典/最新/代表性/对比工作。避免虚假引用、过度引用、无关引用、堆砌、遗漏关键文献、低质量来源、引用与正文不匹配。生成 BibTeX/EndNote/GB-T 7714/APA/IEEE 等格式并按目标 venue 调整。
竞赛与项目申报材料辅助。当用户做统计建模、数学建模、互联网+、挑战杯、大创、创新创业、科研训练等项目时使用。辅助写申报书、项目计划书、商业计划书、路演 PPT、答辩稿、项目摘要、技术路线、创新点、可行性分析、市场分析、研究基础、预期成果、经费预算、团队分工。用于非论文投稿场景,可与论文/软著/专利/PPT 联动。
统一风格与一致性维护。在论文、PPT、图表、代码、项目文档之间保持术语一致、视觉风格一致、逻辑线索一致、创新点表述一致(常驻,所有任务后台生效)。避免同一项目在不同材料中出现说法不一致、指标名称不统一、图表风格混乱、创新点前后矛盾、方法名称变化、数据集名称不统一、论文与 PPT 逻辑不一致、软著与系统功能不一致。
数据处理、数据质量分析与数据集构建。当用户需要清洗数据、处理缺失/异常值、特征工程、数据增强、划分数据集、评估数据质量,或需自建数据集(采集、标注规范、格式、说明文档、隐私合规、发布)时使用。在提 idea 前优先判断现有数据是否足以支撑研究。
Based on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
本仓库把 5 套科研类 Claude Code Skill 包 打包成一个本地 plugin marketplace(插件市场),
安装后即可在 Claude Code 中按需启用。每套包都保留了作者原始的目录结构与相对路径引用,
因此知识库、脚本、模板、_shared/shared/databases 等内部资源都能正确加载。
This repo bundles five academic-research skill packages into a single local Claude Code plugin marketplace. Each package keeps its original directory layout and relative-path references intact, so its knowledge bases, scripts, templates and shared folders all resolve correctly once installed.
| 插件 / Plugin | 技能数 / Skills | 说明 / What it does |
|---|---|---|
| light | 28 | 全流程科研技能包:文献检索、数据工程、创意生成/批判、系统与图表设计、实验分析、论文写作/润色、引用、排版、审稿返修、专利软著、PPT 与竞赛材料。内置 9 个可溯源知识库与可运行脚本。 |
| academic-research-skills | 4 (+35 modes) | 生产级学术研究流水线:research → write → review → revise → finalize,含 deep-research、引用核验闸门与多智能体集成。 |
| nature-skills | 12 | Nature 风格科研技能:学术检索、引用、数据、图表、论文转专利、论文转 PPT、润色、阅读/翻译、审稿回复、同行评审、科研写作,外加 OpenClaw 医学模块。 |
| econ-top-journal-writing | 5 | 经济学顶刊写作流程:总入口路由、英文经济学写作、中文顶刊写作、中英文表图设计、多智能体写作控制器。 |
| writing-ai-paper | 1 | 《Writing AI Conference Papers》新手手册(hzwer & DingXiaoH 著)封装成可调用 skill,用于 AI/ML 顶会论文的选题、框架、引言、可读性与审稿应对。 |
合计 50 个 skill。
前置:已安装 Claude Code。
在 Claude Code 中运行:
# 1) 添加本仓库为插件市场(本地路径,或克隆后的目录)
/plugin marketplace add /home/user/xixi
# 或从 GitHub: /plugin marketplace add caixinyu2017-star/xixi
# 2) 安装需要的插件(市场名为 xixi-research-skills)
/plugin install light@xixi-research-skills
/plugin install academic-research-skills@xixi-research-skills
/plugin install nature-skills@xixi-research-skills
/plugin install econ-top-journal-writing@xixi-research-skills
/plugin install writing-ai-paper@xixi-research-skills
也可以直接运行 /plugin 打开交互式菜单,浏览并勾选要安装的插件。
安装后重启 Claude Code,技能即可被自动触发(也可用 /<skill-name> 显式调用)。
You can also just run /plugin to open the interactive menu, then browse and toggle the plugins you
want. Restart Claude Code after installing; skills then auto-trigger (or invoke explicitly with
/<skill-name>).
部分包附带独立安装脚本(例如 plugins/light/install.sh 会把 28 个 skill 软链到
~/.claude/skills/)。这些脚本仍然可用,但与方式 A 二选一即可,避免重复安装。
Some packages ship their own installer (e.g. plugins/light/install.sh symlinks its 28 skills into
~/.claude/skills/). Those still work, but use either method A or B — not both — to avoid duplicates.
xixi/
├── .claude-plugin/
│ └── marketplace.json # 列出全部 5 个插件 / lists all 5 plugins
├── plugins/
│ ├── light/ # 28 skills (+ databases/, code_assets/)
│ ├── academic-research-skills/ # 4 skills (+ shared/, agents/, hooks/, modes)
│ ├── nature-skills/ # 12 skills (+ skills/_shared/)
│ ├── econ-top-journal-writing/ # 5 skills
│ └── writing-ai-paper/ # 1 skill (handbook wrapped as a skill)
└── README.md
每个插件目录都含有 .claude-plugin/plugin.json 与 skills/。其中 light 与
academic-research-skills 由作者自带插件清单;nature-skills、econ-top-journal-writing
的清单在本次部署中按其结构补全;writing-ai-paper 由原手册文档封装为可调用 skill。
各插件保留了原始 LICENSE/NOTICE:
LICENSE请在使用时遵循各自许可(尤其 academic-research-skills 为非商业 CC-BY-NC-4.0)。
npx claudepluginhub caixinyu2017-star/xixi --plugin lightEconomics top-journal writing workflow: a main routing skill plus general English economics writing (econ-write, synthesized from 50+ authoritative guides), Chinese top-journal writing (cn-top-econ-writing), bilingual regression table & figure design (econ-table-figure-design), and a multi-agent writing controller. A writing/revision system, not a full empirical pipeline.
Writing AI Conference Papers — a beginner's handbook by hzwer (Zhewei Huang) and DingXiaoH (Xiaohan Ding), packaged as an invocable skill. Guidance on finding the core idea, building the framework, writing the introduction/related work, improving readability and defensibility, last-hours checklists, common negative review comments, and which research ideas tend not to work.
Nature-style research skill pack: academic search, citation management, data handling, scientific figures, paper-to-patent, paper-to-PPT, language polishing, reading/translation, reviewer-response drafting, peer review, and Nature-style scientific writing — plus an OpenClaw medical-skills module. Skills share a versioned knowledge layer under skills/_shared.
Production-grade academic research pipeline for Claude Code: research → write → review → revise → finalize. 4 skills, 35+ modes, 38-agent ensemble, v3.7.3 + v3.8 L3 claim-faithfulness gate, v3.9.0 cross-index triangulation, v3.10 triangulation policy layer, v3.11 deterministic citation verification gate (#182).
A growing collection of Claude-compatible academic workflow bundles. Covers scientific figures, manuscript writing and polishing, reviewer assessment, citation retrieval, data availability, paper reading, literature search, response letters, paper-to-PPTX conversion, and evidence-grounded Chinese invention patent drafting. Rules are organized as reusable skill folders with explicit workflows and quality checks.
Complete creative writing suite with 10 specialized agents covering the full writing process: research gathering, character development, story architecture, world-building, dialogue coaching, editing/review, outlining, content strategy, believability auditing, and prose style/voice analysis. Includes genre-specific guides, templates, and quality checklists.
Comprehensive skill pack with 66 specialized skills for full-stack developers: 12 language experts (Python, TypeScript, Go, Rust, C++, Swift, Kotlin, C#, PHP, Java, SQL, JavaScript), 10 backend frameworks, 6 frontend/mobile, plus infrastructure, DevOps, security, and testing. Features progressive disclosure architecture for 50% faster loading.
Build and maintain an LLM-curated personal knowledge base with default local FastEmbed + sqlite-vec search, BM25 fusion, structured evidence, incremental caching, and a compiled graph layer.
UI/UX design intelligence. Searchable local database with 84 styles, 192 palettes, 74 font pairings, 25 charts, and 22 stacks (React, Next.js, Vue, Nuxt.js, Nuxt UI, Svelte, Astro, SwiftUI, React Native, Flutter, Tailwind, shadcn/ui, Jetpack Compose, Angular, Laravel, JavaFX, WPF, WinUI, Avalonia, Uno Platform, UWP, Three.js). Use when designing, building, or reviewing UI: pages, components, color schemes, typography, layout, accessibility, animation, or data visualization.
This skill should be used when users need to generate ideas, explore creative solutions, or systematically brainstorm approaches to problems. Use when users request help with ideation, content planning, product features, marketing campaigns, strategic planning, creative writing, or any task requiring structured idea generation. The skill provides 30+ research-validated prompt patterns across 14 categories with exact templates, success metrics, and domain-specific applications.