Pandas plugins

Programmatically create, edit, and analyze office documents (DOCX, XLSX, PPTX, PDF) and Google Workspace files using Python scripts, enabling automated report generation, format conversion, and data extraction.

Generate and edit images using GPT Image 2 and OpenAI-compatible endpoints with 70+ structured prompt templates across 18 categories, plus build polished visual web artifacts like pages, dashboards, prototypes, slide decks, and data visualizations using HTML/CSS/JavaScript/React.

Run 80+ bioinformatics workflows locally — pharmacogenomics, GWAS, single-cell RNA-seq, ancestry, metagenomics, variant annotation, protein structure prediction, and clinical reporting — with deterministic Python execution, reproducibility bundles, and privacy-preserving local computation.

Guides Chinese academic paper writing from brainstorming through publication: structures research plans, generates literature reviews with BibTeX citations, produces publication-quality Python charts (matplotlib/seaborn), and outputs LaTeX-formatted manuscripts with journal templates.

Build, train, and deploy AI models on AWS SageMaker with deep ML expertise: validate datasets, fine-tune models (SFT, DPO, RLVR), generate Jupyter notebooks, evaluate model quality, and diagnose HyperPod cluster issues (NCCL, GPU, Slurm, EKS) — all from your coding assistant.

Provides 197 computational skills for scientific AI agents to perform life sciences research, covering genomics, proteomics, drug discovery, medical imaging, biostatistics, and scientific writing via integrations with databases, analysis tools, and ML frameworks.

Conduct scientific research across 1000+ biomedical databases (PubMed, UniProt, PubChem, TCGA, FAERS, ClinicalTrials.gov, etc.) through an MCP server, research agent, slash commands, and 115 specialized skills for drug discovery, genomics, clinical trials, and literature review.

Transform WPS Notes into a personal knowledge engine for long-form creative and academic writing. Automates memory retrieval, idea connection, insight generation, structured note-taking from any source (URLs, PDFs, images, audio transcripts), and multi-platform content formatting.

Orchestrate medical research workflows from project intake through publication: design studies, analyze data with reproducible code, audit manuscripts against reporting guidelines, manage references, and generate submission-ready documents and figures.

Develop and backtest quantitative trading strategies for Chinese financial markets (A-shares, futures) using Python frameworks like AKShare, Backtrader, RQAlpha, JoinQuant, QMT, MiniQMT, and Tushare — covering data fetching, event-driven backtesting, parameter optimization, and live trading.

Automates the full academic research workflow: literature search, data processing, idea validation, experimental design, paper drafting and polishing, publication-ready figure generation, peer review simulation, rebuttal crafting, compliance auditing, patent/software registration, and presentation creation, with integrity checks and multi-format typesetting.

Investigate and mitigate Google Cloud incidents using SRE playbooks, anomaly detection on time-series metrics, log analysis, infrastructure discovery, and structured post-mortem generation.

Generate forecasting datasets and fine-tune models using the Lightningrod Python SDK. Create seed data from BigQuery public datasets, local files, or web searches; build SFT and GRPO training data; track experiments; and run transform pipelines for forecasting model development.

Run a structured ML experimentation loop: from EDA and pipeline declaration to evaluation, smoke testing, and audit-driven backlog generation. Use skrub DataOps graphs, skore reports, and a prescribed Python stack to keep experiments reproducible and documented.

Automate end-to-end academic research: write scientific papers with LaTeX/Markdown, search and cite literature, analyze data with Python libraries (pandas, PyTorch), run bioinformatics pipelines, generate publication-quality figures, create posters and presentations, and manage citations and references. Includes tools for grant writing, peer review, and clinical decision support.

Run end-to-end quantitative alpha factor research: validate market data, discover factors via Ralph/Helix loops, evaluate out-of-sample IC/ICIR decay, backtest composite signals under transaction costs, and generate research notes. Connects to financial data APIs (FactSet, Morningstar, etc.) for live data ingestion.

Plan, implement, review, and document the full software development lifecycle — from requirements and architecture to code changes, testing, and release — using a structured skill-based workflow that adapts to the task at hand.

Automate end-to-end ML performance investigations: research SOTA papers and architectures, generate phased plans, judge experimental methodologies, profile bottlenecks, run metric-improvement campaigns with atomic git commits, auto-rollback on regressions, and leverage specialist agents for data lifecycle and deep paper analysis.

Orchestrate an end-to-end academic research workflow inside Claude Code: from literature search and citation verification, through figure design and code-backed implementation, to manuscript drafting, revision, and rebuttal letter assembly — all coordinated by a supervisor agent that tracks bottlenecks and safety gates.

Generate optimized SQL queries from natural language for BigQuery, PostgreSQL, MySQL, and Snowflake; perform cohort analysis on CSV/Excel user data to compute retention rates, visualize trends, and detect anomalies; evaluate A/B tests with statistical significance, confidence intervals, and launch recommendations.

Create and edit Microsoft Office documents (PowerPoint, Word, Excel) programmatically with support for complex formatting, tables, images, styles, and DataFrame exports.

Search arXiv, INSPIRE-HEP, NASA ADS, HEPData, and Zenodo for physics papers and experimental data by keywords, authors, or IDs; download BibTeX, LaTeX source, CSV/YAML/JSON/ROOT files; visualize posterior samples from PolyChord/MultiNest with corner plots, marginals, and KL divergence using matplotlib/pandas.

Automate academic research in finance, economics, and real estate: search literature via Corbis, generate and rank novel ideas with heuristics, screen for novelty and journal fit, visualize trends and gaps with Python-generated figures, draft structured reviews and positioning memos, read/summarize papers with a subagent, and audit citations in LaTeX files.

Automate end-to-end AI/ML academic research workflows: perform topic-driven literature surveys with taxonomies, gaps, and innovations; generate peer/meta-reviews for paper PDFs; detect/insert citations in LaTeX; evaluate/refine ideas into proposals with diagrams/experiments; create reveal.js slides/posters from papers.

Model and analyze electric power networks with pandapower: construct grids from buses, lines, transformers, loads, generators; compute AC/DC power flows, optimal power flow, IEC 60909 short circuits, state estimation, time series simulations, topology checks, and plots.

Streamline end-to-end data science and ML workflows: frame business problems into ML tasks, preprocess and validate data with quality checks, perform EDA on diverse formats, design and execute experiments with hyperparameter tuning via Optuna and interpretability via SHAP, audit reproducibility and leakage, evaluate model performance and readiness for deployment, generate model cards, and extract structured learnings into docs.

Query A-share securities data (K-line, financials, macro, dividends, industry classification) from the free BaoStock API, returning results as pandas DataFrames with documented pitfalls.

Automates end-to-end content creation: collect hot topics from RSS feeds, generate platform-optimized articles for WeChat, Xiaohongshu, and Feishu with AI drafting, enforce compliance review, and publish via automated browser and API workflows. Also supports analyzing content performance data with statistical charts and reports.

Turn any topic into a compact expertise artifact through a structured research pipeline (question tree, source discovery, zero-context fetch, .mv2 indexing, REPL distillation) without loading raw content into LLM context. Execute Python code in isolated Docker sandboxes for data analysis, prototyping, and DSPy sub-agents. Query Neo4j graph databases for knowledge retrieval.