Skill UI

This skill automates the process of vulnerability prioritization by querying the FIRST Exploit Prediction Scoring System (EPSS) API. It calculates the real-world probability of a CVE being exploited within 30 days, providing a crucial risk metric that surpasses traditional CVSS severity scoring. Use it for risk-based vulnerability management, allowing security teams to efficiently triage and remediate the most actively threatened vulnerabilities first.

A foresight agent that runs a rigorous 12-step pipeline combining Claude’s qualitative intelligence with Python’s deterministic arithmetic to analyze high-stakes future scenarios, infer reasonable horizons, and surface quantitative signals for prediction requests.

This skill teaches users to apply the scientific method to software debugging. Instead of random changes, it guides the process through five steps: Observation, Hypothesizing, Prediction, Testing, and Analysis. By systematically forming and testing specific, measurable hypotheses, users can efficiently isolate failing components and confirm the true root cause of technical issues.

A Python library for molecular machine learning, offering diverse featurizers, MoleculeNet benchmarks, graph neural networks, and pretrained models for property and ADMET prediction.

Diffusion-based deep learning tool for predicting 3D binding poses of small-molecule ligands to protein targets from PDB files or sequences plus SMILES/SDF/MOL2. Supports single docking, batch virtual screening, and confidence-score analysis; not for binding affinity prediction.

Use the esm Python SDK with ESM3 and ESMC for protein sequence generation, embeddings, function conditioning, structure prediction, inverse folding, and Forge/Biohub inference workflows.

This skill helps formulate evidence-bounded scientific questions and candidate hypotheses. It turns observations into transparent, testable research plans without treating hypotheses as facts. It distinguishes between observations, mechanisms, and predictions, ensuring rigorous research design and safety compliance.

Build clinical and healthcare deep learning pipelines with PyHealth: load EHR, signal, and imaging datasets, define prediction tasks, instantiate models, train, and evaluate using a modular five-stage workflow.

Build, evaluate, and audit right-censored or competing-risk survival workflows with scikit-survival 0.28.0, covering leakage-safe pipelines, model selection, time-dependent prediction, and censoring-aware metrics such as IPCW concordance, dynamic AUC, and Brier scores.

This skill covers using the SHAP library to explain and audit machine learning predictions. It guides users on selecting explainers, computing feature attributions, and validating results. It includes workflows for local and global visualizations, handling multi-output models, and ensuring model behavior descriptions are accurate without implying causality. Requires Python 3.12+ and specific SHAP versions.

Utilize Google's TimesFM foundation model for zero-shot univariate time series forecasting without training custom models. Supports sales, sensor, and weather data with point forecasts and prediction intervals. Includes mandatory system checks for RAM and GPU before inference to prevent crashes.

TorchDrug is a modular PyTorch library for graph learning tasks including molecular property prediction, molecule generation, and knowledge graph reasoning. It provides datasets, models, and training engines for building workflows in computational chemistry and biology. Ensure compatibility with PyTorch 1.8-2.0 and Python 3.7-3.10 before implementation.