Skill UI

This skill provides expert capabilities for training and fine-tuning transformer language models using the TRL (Transformers Reinforcement Learning) library. It supports state-of-the-art post-training methods, including Supervised Fine-Tuning (SFT), Direct Preference Optimization (DPO), Group Relative Policy Optimization (GRPO), and Reward Model training. Use this skill to align and customize foundation models for advanced tasks via CLI commands.

A comprehensive guide covering scalable backend architecture patterns, including Repository, Service, and Middleware design. It details best practices for building robust APIs using Node.js, Express, and Next.js. Topics covered include RESTful API structuring, database query optimization (preventing N+1 issues), implementing database transactions, and integrating caching strategies like Redis. This guide is essential for developing high-performance, maintainable server-side applications.

This skill converts vague speed requests into a structured optimization process. It establishes a baseline, defines correctness gates, and iteratively tests variants against metrics like latency or cost. Each variant is benchmarked and validated before promotion, ensuring reproducible improvements without breaking functionality. Ideal for performance tuning, recursive optimization, and selecting the best implementation through measured tests.

A comprehensive guide to high-performance analytical database patterns using ClickHouse. Covers advanced topics including MergeTree engine types (Replacing, Aggregating), efficient query optimization techniques (filtering, window functions), bulk data insertion strategies, and setting up materialized views for real-time analytics. Ideal for data engineers and analysts building OLAP systems.

A comprehensive guide to building highly performant and reusable UIs across multiple platforms (Android, iOS, Desktop, Web) using Compose Multiplatform and Jetpack Compose. It details advanced patterns for state management (Single State Object, Event Sink), type-safe navigation, slot-based composables, platform adaptation (Expect/Actual), and crucial performance optimization techniques.

This tool audits the token consumption across all components (agents, skills, rules, MCP servers) within a Claude Code session. It analyzes token overhead, identifies bloat (e.g., overly long descriptions, redundant components), and generates a detailed context budget report. Use it to maintain optimal context window usage, improve performance, and prevent context saturation.

A comprehensive, library-agnostic checklist covering best practices for Flutter and Dart applications. It guides developers on architectural principles, state management patterns (BLoC, Riverpod, etc.), Dart idioms, performance optimization, and clean code principles, making it invaluable for code reviews.

A comprehensive guide detailing modern architectural patterns for building high-performance user interfaces with React and Next.js. Covers advanced techniques including Compound Components, Render Props, custom state hooks (like useQuery and useDebounce), advanced state management (Context + Reducer), and performance optimization strategies like memoization and code splitting.

This tool provides a structured way to run predefined, periodic background workers (e.g., security audits, performance optimization, memory consolidation). It utilizes native scheduling mechanisms to automatically dispatch and re-schedule tasks within a session, ensuring continuous system maintenance and optimal performance monitoring.

A comprehensive guide detailing advanced performance optimization patterns for React 18/19 and Next.js. Adapted from Vercel Engineering's best practices, this skill organizes 70+ critical rules across eight categories—including eliminating waterfalls, optimizing bundle size, improving server-side rendering, and managing client-side re-renders. Use this when developing, reviewing, or refactoring high-performance React applications.

A comprehensive guide to building high-performance, scalable UIs using the latest SwiftUI best practices. Covers advanced state management with @Observable, type-safe navigation using NavigationStack, view composition techniques, and crucial performance optimization tips for iOS and macOS development.

This skill performs mean-variance portfolio optimization by solving the linear system Σ · x = μ using the Conjugate Gradient (CG) method. It is significantly faster (40-60x) and numerically robust compared to legacy methods like the Neumann series. It takes a covariance matrix (Σ) and expected return vector (μ) to determine optimal asset weights, providing full provenance tracking for auditability.