Machine Learning

optimize-for-gpu helps turn CPU-bound Python into NVIDIA GPU code with the right library choice. Use it for arrays, dataframes, ML pipelines, graph analytics, imaging, geospatial work, vector search, and custom kernels. It guides CuPy, cuDF, cuML, cuGraph, cuCIM, cuVS, KvikIO, Numba CUDA, and Warp decisions with practical optimize-for-gpu usage and migration advice.

hypogenic is a skill for generating and testing hypotheses on tabular or text-derived datasets with LLM support. It helps with hypogenic for Data Analysis by turning empirical questions into structured, testable workflows for classification interpretation, content analysis, and deception detection. Use it when you need evidence-backed hypotheses, not just brainstorming.

diffdock is a docking skill for predicting protein-ligand binding poses from PDB structures or protein sequences plus ligands in SMILES, SDF, or MOL2. Use the diffdock skill for structure-based drug design, virtual screening, and confidence-scored pose analysis. It is not for binding affinity prediction.

pytdc is a skill for Therapeutics Data Commons, giving AI-ready drug discovery datasets and benchmarks for ADME, toxicity, DTI, DDI, generation, scaffold splits, and pharmacological prediction.

pytorch-lightning skill for organizing PyTorch projects with LightningModules and Trainers. Use this pytorch-lightning guide for install, training, validation, logging, checkpointing, and distributed execution across multi-GPU or TPU workflows.

pymoo is a Python skill for single- and multi-objective optimization, Pareto fronts, constrained problems, and benchmark tests. Use this pymoo guide to choose algorithms like NSGA-II, NSGA-III, and MOEA/D, follow the install and usage workflow, and apply pymoo for Data Analysis when multiple metrics must be balanced.

pyhealth helps you build clinical and healthcare deep-learning pipelines with a Dataset → Task → Model → Trainer → Metrics workflow. Use this pyhealth skill for MIMIC-III/IV, eICU, OMOP, SleepEDF, ChestXray14, EHRShot, prediction, drug recommendation, sleep staging, ICD coding, EEG events, and medical code mapping.

pufferlib is a high-performance reinforcement learning skill for fast parallel simulation, vectorized rollouts, and multi-agent training. Use this pufferlib guide to install, understand pufferlib usage, and adapt RL pipelines with Gymnasium, PettingZoo, Atari, Procgen, or NetHack-style environments. Ideal for code generation focused on throughput and scalable PPO workflows.

molfeat is a molecular featurization skill for ML and Data Analysis. It helps convert SMILES or RDKit molecules into fingerprints, descriptors, and pretrained embeddings for QSAR, virtual screening, similarity search, and chemical space analysis. Use this molfeat guide to pick practical representations and build reusable featurization pipelines.

geniml is a skill for genomic interval machine learning on BED files, scATAC-seq outputs, and chromatin accessibility data. Use it for Region2Vec, BEDspace, scEmbed, consensus peaks, and other region-level ML workflows. It is a good fit when you need embeddings, clustering, or preprocessing guidance for genomic regions.

esm skill for protein language models, including ESM3 generation and ESM C embeddings. Use this esm guide for protein sequence design, inverse folding, function prediction, and code generation workflows with local inference or the Forge API.

cellxgene-census skill for querying the CELLxGENE Census programmatically. Use it to explore expression data, metadata, embeddings, and cross-dataset patterns across tissues, diseases, and cell types. Best for population-scale single-cell analysis and reference atlas comparisons; for your own data, use scanpy or scvi-tools.

aeon is a scikit-learn-compatible Python skill for time series machine learning. Use it for classification, regression, clustering, forecasting, anomaly detection, segmentation, similarity search, and other temporal data workflows. It fits univariate and multivariate analysis when you need specialized methods beyond generic tabular ML.

detecting-deepfake-audio-in-vishing-attacks helps security teams analyze audio for AI-generated speech in vishing, fraud, and impersonation cases. It extracts spectral and MFCC-based features, scores suspicious samples, and produces a forensic-style report for review. Ideal for Security Audit and incident response workflows.

Detect business email compromise with AI using NLP, stylometry, behavioral signals, and relationship context. This detecting-business-email-compromise-with-ai skill helps SOC, fraud, and Security Audit teams score suspicious emails, explain risk signals, and decide whether to quarantine, warn, or escalate.