Data Science & ML Claude 技能

Datamol Cheminformatics

Simplifies molecular cheminformatics and drug discovery workflows using a Pythonic interface for RDKit.

Biomni: Biomedical AI Agent

Executes autonomous multi-step biomedical research tasks across genomics, drug discovery, and clinical analysis using integrated databases and code execution.

Geniml Genomic Machine Learning

Applies unsupervised machine learning models to genomic interval data for embedding, clustering, and bioinformatics analysis.

Model Equivariance Auditor

Verifies that machine learning models correctly respect intended symmetries through systematic numerical tests and debugging guidance.

Forecast Pre-Mortem

Stress-tests predictions by assuming failure and working backward to identify hidden risks and blind spots.

PDB Database Access

Accesses the RCSB Protein Data Bank to search, retrieve, and analyze 3D structural data for biological macromolecules.

Health Science & Clinical Research

Structures clinical research and evaluates health evidence using established evidence-based medicine frameworks like PICOT and GRADE.

Causal Inference & Root Cause Analysis

Systematically investigates causal relationships to identify fundamental root causes and distinguish them from symptoms or correlations.

Code Data Analysis Scaffolds

Provides structured frameworks and systematic workflows for Test-Driven Development (TDD), data exploration, and statistical modeling.

Seaborn Statistical Visualization

Generates publication-quality statistical graphics and exploratory data visualizations using the Seaborn Python library.

Bayesian Reasoning & Calibration

Implements systematic Bayesian probability updates to improve forecasting and decision-making under uncertainty.

SHAP Model Interpretability

Explains machine learning model predictions and feature importance using SHapley Additive exPlanations (SHAP) across various model architectures.

Reference Class Forecasting

Establishes objective statistical baselines for predictions and projects by anchoring them in historical reality and the 'Outside View.'

D3.js Data Visualization

Builds custom, interactive data-driven visualizations using D3.js for complex layouts, maps, and bespoke charting requirements.

Symmetry Group Identifier

Maps identified symmetries to mathematical groups to formalize requirements for equivariant and invariant neural network architectures.

Strategic Decision & Business Case Analyst

Quantifies uncertain business choices and crafts persuasive, data-backed narratives to secure stakeholder buy-in.

Visualization Choice & Reporting

Matches data questions to optimal chart types and generates narrated reports that turn raw metrics into actionable business insights.

Equivariant Architecture Designer

Designs neural network architectures that preserve symmetry groups to improve model efficiency and robustness.

FlowIO Cytometry Handler

Parses and creates Flow Cytometry Standard (FCS) files for seamless integration with Python data science workflows.

Aeon Time Series ML

Simplifies complex time series machine learning tasks including forecasting, classification, and anomaly detection using a scikit-learn compatible toolkit.

Scout Mindset Bias Check

Detects and mitigates cognitive biases to improve decision-making, forecasting accuracy, and intellectual honesty.

Glide AI Integration

Enhances Glide apps with AI-powered columns for text generation, OCR, transcription, and automated data processing.

Adaptive Walk-Forward Epoch Selection

Optimizes training epochs for machine learning models using Walk-Forward Efficiency and efficient frontier analysis.

Range Bar Evaluation Metrics

Evaluates financial models using state-of-the-art metrics specifically designed for range bar (price-based sampling) data.

Replicate CLI

Interacts with the Replicate AI platform to run models, manage deployments, and perform fine-tuning directly from the command line.

Advanced Excel & Spreadsheet Toolkit

Automates professional spreadsheet creation, complex financial modeling, and data analysis with formula-driven accuracy.

Council Consultation Ensemble

Leverages a multi-model ensemble to provide diverse, parallel insights from GPT, Gemini, and Claude for complex problem-solving.

Prompt Architecture Reviewer

Analyzes and evaluates LLM prompts using a structured 10-Layer Architecture to improve clarity, precision, and output performance.

Expected Value

Calculates probability-weighted outcomes to facilitate rational decision-making under uncertainty for investments, product development, and strategic initiatives.

Scientific Hypothesis Generation

Generates evidence-based, testable scientific hypotheses and structured experimental designs to accelerate research and discovery.