Skill UI

An expert skill for transforming CPU-bound Python code into highly performant GPU-accelerated code using NVIDIA technologies. It specializes in optimizing scientific workflows, including machine learning (ML), physics simulations, geospatial analysis, and large-scale array computations. Utilize tools like CuPy for array operations, Numba CUDA for custom kernels, and Warp for spatial computing and differentiable simulations, achieving dramatic speedups (10x to 1000x). Use this when dealing with NumPy, pandas, SciPy, or complex computational loops that benefit from parallel processing on NVI...

A versatile, hardware-agnostic quantum computing framework designed for training quantum circuits and building hybrid quantum-classical models. It provides automatic differentiation, supporting advanced applications in Quantum Machine Learning, Quantum Chemistry (VQE), and variational algorithms, while ensuring portability across major quantum hardware backends (IBM, Google, etc.).

Polars is a lightning-fast DataFrame library built on Apache Arrow for Python. It provides an expression-based API, lazy query optimization, and parallel computation capabilities for efficient data manipulation, analytics, and ETL pipelines. It is designed to handle large, memory-intensive datasets, making it an ideal alternative for migrating from pandas or optimizing complex data workflows.

PufferLib is a high-performance, scalable RL framework designed for massive parallel training and vectorized simulation. Featuring PuffeRL (optimized PPO+LSTM) and the PufferEnv API, it supports multi-agent systems, custom environment creation, and achieves industry-leading speeds (up to millions of steps/second). Ideal for advanced research and large-scale AI experimentation.

PyTDC (Therapeutics Data Commons) is an open-science platform providing comprehensive, AI-ready datasets and benchmarks for the entire drug discovery and development pipeline. It supports single-instance prediction (e.g., ADME, toxicity), multi-instance prediction (DTI, DDI, PPI), and novel molecule generation (MolGen). Use this skill for benchmarking machine learning models and predicting complex biomedical properties.

Pyzotero is a powerful Python wrapper for the Zotero Web API v3. It enables developers to programmatically interact with Zotero libraries, allowing for CRUD operations on items, collections, tags, and attachments. Use this skill to build robust research automation workflows, manage bibliographic references, export citations (e.g., BibTeX), or perform bulk data processing within a Zotero environment.

Qiskit is a leading open-source framework for quantum computing. It enables users to build, optimize, and execute complex quantum circuits and algorithms. The platform supports running on various hardware backends, including IBM Quantum, IonQ, and local simulators. Ideal for advanced research in quantum chemistry simulations, variational quantum algorithms (VQE/QAOA), and quantum machine learning.

A comprehensive resource detailing the use of scikit-learn, the industry-standard Python library for classical machine learning. Use this skill for the entire ML lifecycle, including supervised tasks (classification, regression), unsupervised tasks (clustering, dimensionality reduction), data preprocessing, model evaluation, and building production-ready pipelines.

SHAP (SHapley Additive exPlanations) provides a unified, robust framework for understanding complex machine learning models. It calculates feature importance using Shapley values from game theory, allowing users to interpret individual predictions, analyze model bias, visualize feature interactions, and debug model behavior. Applicable across deep learning, tree-based, and black-box models.

UMAP (Uniform Manifold Approximation and Projection) is a powerful technique for non-linear dimensionality reduction. This skill enables the creation of scalable, structure-preserving embeddings (2D/3D) for visualization, clustering preprocessing, and advanced machine learning feature engineering. It supports various metrics (e.g., Euclidean, Cosine) and supervised learning workflows.

Vaex is a high-performance Python library designed for processing and visualizing enormous tabular datasets that exceed available RAM (gigabytes to terabytes). It excels at out-of-core DataFrame operations, lazy evaluation, fast statistical aggregation, and machine learning integration on massive datasets. Use it for analyzing large CSV, Parquet, or HDF5 files efficiently on a single machine, without memory constraints.

This advanced skill optimizes LLM context usage by implementing strict rules for prompt caching, efficient log pruning, AST-based code inspection, and compact data structuring. It minimizes token consumption and maximizes information density, suitable for highly technical and iterative development workflows.