Data Science & ML Claude 技能

Hyperspell Setup

Integrates Hyperspell’s long-term memory and RAG capabilities into your project with automated configuration and SDK setup.

Mozilla Llamafile Integration

Configures and manages local LLM inference using Mozilla Llamafile to provide offline, OpenAI-compatible AI capabilities.

Prompt Engineering Expert

Optimizes AI prompts using research-backed frameworks and production-ready templates to ensure high-quality, cost-effective model outputs.

LiteLLM Unified Interface

Simplifies LLM integration by providing a consistent OpenAI-style Python interface for over 100 cloud and local providers.

Llamafile Local LLM

Configures and manages local LLM inference using Mozilla's llamafile to provide offline, OpenAI-compatible AI capabilities.

Excel & Financial Modeling

Creates, analyzes, and manages complex Excel spreadsheets with a focus on financial modeling standards and formula integrity.

Qualitative Research Specialist

Guides the design and execution of rigorous qualitative studies, thematic analysis, and coding schemes for non-numerical data.

Quantitative Analysis & Statistics

Conducts rigorous statistical analysis, power calculations, and data interpretation using PhD-level methodologies.

Lateral Thinking Specialist

Facilitates breakthrough problem-solving by applying first-principles reasoning and cross-domain analogies to overcome complex technical bottlenecks.

Methodology Design Specialist

Provides PhD-level guidance for selecting, designing, and validating academic and scientific research methodologies.

Co-Researcher Orchestrator

Facilitates PhD-level academic research by managing specialized methodology skills and enforcing rigorous scientific standards.

Python Performance Optimizer

Profiles and optimizes Python code to identify bottlenecks, reduce latency, and minimize memory consumption using industry-standard tools.

LLM Task Delegation Core

Orchestrates task delegation to external LLM services by offloading high-token execution while maintaining central reasoning within Claude.

AI Research Analyst

Conducts deep-dive information gathering and data synthesis to produce actionable strategic insights and comprehensive reports.

LLM Architect

Designs, optimizes, and deploys scalable large language model architectures and high-performance RAG systems.

Prompt Engineering Expert

Optimizes LLM output quality by providing domain-specific guidance and structural patterns for crafting high-performance prompts.

Hugging Face Transformers

Loads and deploys state-of-the-art pretrained models for text, vision, and audio using the Hugging Face library.

Apple Silicon MLX Engineering

Optimizes and runs large language models on Apple Silicon using the native MLX framework for high-performance inference and fine-tuning.

AI Agent & Workflow Patterns

Architects sophisticated AI agent systems and LLM workflows using proven patterns like ReAct, prompt chaining, and orchestrator-worker models.

RLHF Alignment Guide

Provides comprehensive technical guidance on Reinforcement Learning from Human Feedback for aligning large language models with human preferences.

Research Methodology Expert

Matches complex research questions to rigorous experimental designs, sampling strategies, and validity controls using PhD-level frameworks.

Context Degradation Diagnostics

Identifies and mitigates performance loss in large-context agent systems using proven architectural patterns and empirical benchmarks.

Context Engineering Fundamentals

Masters the mechanics of LLM context to design efficient, high-performance agent architectures and debugging strategies.

Context Optimization for AI Agents

Optimizes AI agent context windows using advanced compaction, masking, and partitioning techniques to improve performance and reduce token costs.

Memory Systems for AI Agents

Design and implement advanced memory architectures to help AI agents persist state, maintain entity consistency, and reason over structured knowledge.

Multi-Agent Architecture Patterns

Designs and implements robust multi-agent systems to overcome context limits and handle complex, parallelizable tasks.

Context Degradation Diagnostic & Mitigation

Diagnoses and mitigates performance loss in AI agents by identifying patterns like lost-in-middle, context poisoning, and distraction.

Multi-Agent Architecture Patterns

Designs and implements scalable multi-agent systems to overcome context limitations and manage complex task decomposition.

Context Engineering Fundamentals

Master the mechanics, constraints, and optimization strategies of context within AI agent architectures to improve performance and reduce costs.

Advanced AI Evaluation & LLM-as-a-Judge

Implements robust LLM-as-a-Judge evaluation techniques to measure, compare, and optimize the quality of AI-generated outputs.