Data Science & ML Claudeスキル

Scikit-HEP Hist Specialist

Streamlines the creation, manipulation, and visualization of multidimensional histograms using the scikit-hep Python ecosystem.

MechInterp Feature Overview

Provides a rapid diagnostic summary of Sparse Autoencoder (SAE) features to generate research hypotheses and identify model behaviors.

Financial Change-Point Detection

Detects structural breaks and regime shifts in financial time-series using Gaussian Process models to identify market transitions.

Notebook Creator

Generates standardized Jupyter notebooks for fantasy football data analysis, integrating DuckDB connections, dbt mart queries, and professional visualization patterns.

AI Text Classifier & ML Engine

Categorizes text, detects sentiment, and filters spam using pre-trained and custom-trained machine learning models.

Document Ranking and Scoring

Ranks and filters retrieved documents based on vector similarity metrics to optimize RAG pipeline relevance.

Sprint 1 Executor

Executes structured, atomic tasks for fantasy football analytics, FASA optimization, and trade intelligence within a standardized sprint framework.

CarsXE VIN OCR

Extracts Vehicle Identification Numbers (VIN) from images and photos using advanced OCR technology.

Transformer Architecture Guide

Explains and implements core Transformer architecture components for LLM development, fine-tuning, and model analysis.

LLM Evaluation & Optimization

Evaluates LLM outputs and optimizes prompts using Evidently.ai metrics and LLM-as-a-judge patterns.

Parameter-Efficient Fine-Tuning (PEFT)

Optimizes LLM fine-tuning using LoRA, QLoRA, and Unsloth to drastically reduce memory requirements and accelerate training cycles.

Ollama Python Library

Provides a clean, Pythonic interface for interacting with Ollama to handle text generation, chat completions, and model management.

Direct Preference Optimization (DPO)

Aligns AI models with human preferences using Direct Preference Optimization to improve reasoning and response quality without explicit reward models.

Quantization & Model Optimization

Optimizes large language models for efficient inference and training using various precision types and memory estimation techniques.

LLM Supervised Fine-Tuning (SFT)

Streamlines the supervised fine-tuning of Large Language Models using Unsloth for optimized performance and reasoning model development.

LangChain Framework Support

Simplifies building LLM-powered applications by providing standardized abstractions for prompt engineering, model orchestration, and structured output parsing.

Ollama Local LLM Manager

Manages local Ollama inference servers using Podman Quadlet to provide GPU-accelerated LLM capabilities.

ComfyUI Manager

Deploys and manages ComfyUI instances for node-based Stable Diffusion image generation with GPU acceleration and model lifecycle support.

RLOO Policy Optimization

Implements Reinforcement Learning with Leave-One-Out estimation to stabilize model training and optimize policy performance.

Bayesian Meta-Analysis Models

Provides comprehensive Bayesian meta-analysis templates using Stan and JAGS for advanced biostatistical evidence synthesis.

Marimo Notebook Assistant

Streamlines the creation and management of reactive marimo notebooks for interactive data science and analytics workflows.

Marimo Notebook Inspector

Inspects Marimo notebook execution results and HTML snapshots to debug errors and verify cell outputs.

Marimo Development Server

Deploys and manages reactive Python notebooks with hot-reloading capabilities for interactive development.

Demographic Predictor

Predicts age, gender, and ethnicity from person data and images to enrich datasets and customer profiles.

Workers AI Expert

Deploys and optimizes serverless AI models, embedding generation, and RAG architectures directly on Cloudflare’s edge network.

Meta-Analysis Master (C5)

Orchestrates comprehensive meta-analysis workflows with multi-gate validation to ensure data integrity and statistical accuracy.

Gemini Video Intelligence

Analyzes and extracts deep insights from video files and YouTube URLs using the Google Gemini API.

Spark Optimization

Optimizes Apache Spark data processing jobs through advanced partitioning, memory management, and shuffle tuning.

Trading Model Training Gating

Automates the archival and quality classification of algorithmic trading models based on performance metrics and risk thresholds.

I3 RAG Builder & Document Processor

Builds cost-free RAG systems using parallel document processing and local vector embeddings.