ai-agent-research-starter-kit

Convert files and office documents to Markdown. Supports PDF, DOCX, PPTX, XLSX, images (with OCR), audio (with transcription), HTML, CSV, JSON, XML, ZIP, YouTube URLs, EPubs and more.

Use this skill when users need to search academic papers, download research documents, extract citations, or gather scholarly information. Triggers include: requests to "find papers on", "search research about", "download academic articles", "get citations for", or any request involving academic databases like arXiv, PubMed, Semantic Scholar, or Google Scholar. Also use for literature reviews, bibliography generation, and research discovery. Requires OpenClawCLI installation from clawhub.ai.

Search academic papers and conduct literature reviews using OpenAlex API (free, no key needed). Use when the user needs to find scholarly papers by topic/author/DOI, explore citation chains, get structured paper metadata (title, authors, abstract, citations, DOI, open access URL), fetch full text of open access papers, or conduct automated literature reviews with theme identification and synthesis. Triggers on requests involving academic search, paper lookup, citation analysis, literature review, research synthesis, or scholarly reference gathering.

Use AI4Scholar author tools to search scholars, inspect author profiles, retrieve an author's papers, compare experts, identify potential reviewers, map labs or collaborators, and verify whether an author is the correct person.

Use AI4Scholar auto_cite to add real citations to academic text and return formatted references and BibTeX. Trigger for automatic citation insertion, APA/IEEE/Vancouver/Nature citation support, reference generation, BibTeX export, or checking whether claims have real supporting papers.

Use AI4Scholar to trace citation networks: citing papers, references, PubMed related papers, backward/forward citation search, classic paper discovery, mechanism literature expansion, and reviewer-style citation gap checks.

Use AI4Scholar to retrieve and verify detailed metadata for one paper or many papers by DOI, PMID, arXiv ID, Semantic Scholar ID, or title. Trigger for paper detail lookup, batch paper metadata, DOI verification, PMID verification, BibTeX preparation, reference cleanup, or checking whether a cited paper is real.

Use AI4Scholar recommendation tools to find related papers from one or more seed papers, build reading lists, expand a literature map, and discover papers adjacent to a user's theory, method, dataset, or empirical context.

Use AI4Scholar to search real scholarly papers across Semantic Scholar, PubMed, Google Scholar, arXiv, bioRxiv, and medRxiv. Trigger for AI4Scholar paper search, latest literature, top journal literature discovery, cross-database search, paper search prompts, or when the user needs traceable real references rather than invented citations.

Use AI4Scholar to download open-access PDFs or read full text from Semantic Scholar, arXiv, bioRxiv, medRxiv, and DOI-based sources when access permits. Trigger for reading papers, extracting methods/results, PDF download, full-text summaries, or DOI full-text retrieval.

Use AI4Scholar for real scholarly literature tasks through MCP or the OpenClaw plugin: paper search, Google Scholar/Semantic Scholar/PubMed/arXiv/bioRxiv/medRxiv queries, PDF download and reading, citation network lookup, author lookup, paper recommendations, auto_cite for real references, BibTeX, and sci_draw scientific figures. Trigger when the user mentions AI4Scholar, ai4scholar.net, mcp.ai4scholar.net, auto citation, real references, OpenClaw scholar plugin, Scholar Mode, or wants traceable citations.

Use AMiner MCP for academic knowledge graph tasks: scholar search, author profile lookup, institution or team analysis, paper and patent discovery, academic influence checks, and research trend mapping. Trigger when the user mentions AMiner, AMiner MCP, mcp.aminer.cn, scholar portrait, academic graph, institution comparison, expert discovery, patent search, or wants to combine AMiner with literature review and citation verification.

Search and summarize papers from ArXiv. Use when the user asks for the latest research, specific topics on ArXiv, or a daily summary of AI papers.

Comprehensive citation management for academic research. Search Google Scholar and PubMed for papers, extract accurate metadata, validate citations, and generate properly formatted BibTeX entries. This skill should be used when you need to find papers, verify citation information, convert DOIs to BibTeX, or ensure reference accuracy in scientific writing.

Conduct comprehensive, systematic literature reviews using multiple academic databases (PubMed, arXiv, bioRxiv, Semantic Scholar, etc.). This skill should be used when conducting systematic literature reviews, meta-analyses, research synthesis, or comprehensive literature searches across biomedical, scientific, and technical domains. Creates professionally formatted markdown documents and PDFs with verified citations in multiple citation styles (APA, Nature, Vancouver, etc.).

Build high-quality literature reviews from a research topic using a 10-phase workflow. Use when the user asks for 文献综述，文献回顾，literature review, literature survey, 帮我找文献，or 中英文文献整理. ⚠️ **依赖声明:** 本技能依赖以下外部组件，使用前请确保已安装: - cnki-crawler (CNKI 爬虫技能) - academic-research (OpenAlex 检索技能) - academic-research-hub (Google Scholar 检索技能) - PostgreSQL 数据库 (cnki_db) 📦 **首次使用:** 请先阅读 `INSTALL.md` 并运行 `python scripts/check_dependencies.py`

批量提取PDF论文的创新点并生成结构化markdown文档；当用户需要分析学术论文、总结研究贡献或整理技术要点时使用

从PDF学术论文中自动提取结构化信息（标题、作者、单位、期刊信息、研究问题、方法、结论），生成Markdown格式的论文总结文档；适用于需要快速梳理多篇论文核心内容的场景，如文献综述、学术研究或知识管理

Look up current research information using the Parallel Chat API (primary) or Perplexity sonar-pro-search (academic paper searches). Automatically routes queries to the best backend. Use for finding papers, gathering research data, and verifying scientific information.

Introduction to literature search & review skills - systematic paper finding, screening, extraction, and citation traversal

基于Phillip Chong Ho Shon方法论系统解码SSCI论文各部分，提供结构化剖析与批判性解读；当用户需要分析学术论文、撰写文献综述、理解论文特定部分（摘要、引言、方法、结果、讨论等）或进行文献对话时使用

Crawl CNKI journal paper metadata with the local Python crawler and save to PostgreSQL. Supports multi-sort retrieval (DFR/CF/PT/ZH) and mandatory query validation. Use when the user asks to collect CNKI papers, run the CNKI crawler, or query by CNKI field patterns.

CNKI（中国知网）高级搜索自动化技能。使用浏览器自动化技术搜索文献并获取结果列表及摘要信息。建议在有头浏览器环境下使用以便于处理反机器人验证。

CNKI 选题分析助手 - 整合热榜追踪、趋势分析、文献调研、期刊匹配和 6 维度研究空白分析，支持一键自动化分析，为学术选题提供数据驱动的决策支持。

Crawl CNKI `getGroupData` for a specific keyword, keep only non-empty group results, and generate a trend-analysis report with charts focused on the current year, recent yearly momentum, topic concentration, subject and journal concentration, author and institution distribution, and fund signals. Use when the user asks for a CNKI keyword trend report, a 知网关键词趋势分析, a 年度趋势汇报, or wants CNKI data fetched first and then interpreted.

对CSV/Excel数据进行全面的统计描述、相关性分析和可视化；当用户需要数据分析、数据探索、统计摘要、特征分布分析或相关性分析时使用

对CSV或Excel数据进行全面的可视化分析，包括描述性统计、相关性分析、特征分布图和热力图，并生成HTML格式的交互式分析报告；当用户需要数据分析、数据可视化、探索性数据分析或生成数据报告时使用

提供决策树分类建模与超参数优化能力；当用户需要建立分类模型、优化模型参数、生成可视化决策树或输出完整建模报告时使用

Parameterized Python empirical-analysis and machine-learning workflow for applied economics, public health epidemiology, supervised ML, and ML causal inference. Use when the user asks for data cleaning, feature engineering, train/test/validation splits, feature matrix X and target y, Table 1, diagnostic tests, OLS/panel/IV-style formulas, LinearRegression, Ridge, Lasso, ElasticNet, decision trees, random forests, GBDT, regression/classification metrics, DID/event-study formulas, DML/double machine learning with LinearDML, causal forests, robustness checks, mechanism or heterogeneity analysis, mediation, publication-ready tables, figures, or an end-to-end empirical paper pipeline. The skill must route execution through fixed step scripts under scripts/ instead of writing ad hoc Python code in markdown.

Automatically merge scattered Excel and CSV files, normalize column names, and extract structured tables from PDF, HTML, TXT, or Markdown documents. Supports OCR for scanned PDFs. Use when the user wants 多源数据整合、Excel/CSV 自动合并、PDF 表格提取、年报表格提取、政策文件表格抽取、OCR 文字识别，or needs a directory of mixed tabular sources transformed into consolidated CSV outputs.

Automatically analyze research datasets and generate publication-ready statistical tables, correlation figures, grouped comparison plots, and regression diagnostics. Use when the user wants科研数据自动分析、学术绘图、实证数据可视化、可发表级图表输出, or needs a CSV/Excel dataset turned into summary statistics and clean PNG figures.

获取中国金融市场数据（A股、港股、美股、基金、期货、债券）。支持220+个Tushare Pro接口：股票行情、财务报表、宏观经济指标。当用户请求股价数据、财务分析、指数行情、GDP/CPI等宏观数据时使用。

Route text-labeling requests across LDA topic modeling, sklearn baselines, pretrained transformer models, and OpenAI-compatible LLM labeling. First inspect the dataset and infer columns/defaults automatically, then wait for explicit user confirmation before execution.

Paddle-based deep learning workflows from the course materials, including DNN/RNN text-style baselines and the CNN/LeNet image classification case using folder-labeled digit images. Use when a user asks Codex to train, evaluate, save, load, or predict with course DNN, CNN/LeNet, or RNN/GRU examples.

Chinese sentiment analysis workflows using dictionary scoring, traditional machine learning classifiers, and LLM API calls. Use when a user asks Codex to label sentiment, calculate dictionary sentiment scores, train a TF-IDF sentiment classifier, predict sentiment from a saved classifier, or call a large language model for sentiment classification on text files or short inputs.

Basic Chinese NLP and text analysis workflows for PDF text/table extraction, jieba tokenization, word and sentence frequency, word clouds, TF-IDF, Word2Vec, sentence embeddings, and text similarity. Use when a user asks Codex to extract text from PDFs, segment Chinese text, count terms or sentences, create word clouds, vectorize text, train Word2Vec, encode embeddings, or compute text similarity from files or short inputs.

Topic modeling workflows for Chinese corpora using LDA, Dynamic Topic Model/DTM, and BERTopic. Use when a user asks Codex to discover topics, train LDA models, compute topic keywords, model topic evolution over time, train DTM, run BERTopic, export document-topic assignments, or create topic visualization outputs from text files or tabular corpora.

依据实证论文的规范结构与写作风格生成学术论文；适用于生成金融、经济、会计等领域的研究论文框架与正文内容

Generate testable hypotheses. Formulate from observations, design experiments, explore competing explanations, develop predictions, propose mechanisms, for scientific inquiry across domains.

Automatically consolidate literature metadata, generate a paper outline, normalize citation keys, and output LaTeX, Word, and PDF manuscript drafts.