Skill UI

Enables Claude to configure neural network architectures through the neural-network-builder plugin, creating or modifying models and defining layers, activation functions, and training parameters whenever network design or tuning is requested.

Provides expert guidance on designing, implementing, and optimizing cutting-edge computer vision pipelines. Expertise spans real-time object detection (YOLO26), advanced promptable segmentation (SAM 3), and leveraging Vision Language Models (VLMs) for complex visual reasoning. Ideal for building high-performance systems in robotics, autonomous vehicles, and industrial inspection, including 3D reconstruction and edge deployment optimization.

Guides setting up Constitutional AI by combining self-critique SL with RLAIF training to align generative models without human labels and reduce harmful outputs in systems like Claude.

This skill analyzes the predictable ways language models lose performance as context length increases. It details core patterns, including 'lost-in-middle,' context poisoning, and distraction. Use this knowledge to diagnose unpredictable agent failures, design highly reliable systems for long document processing, and implement advanced context engineering strategies.

A comprehensive guide to understanding the complete state (context) available to Large Language Models during inference. Covers core components like System Prompts, Tool Definitions, Message History, and Retrieval Augmented Generation (RAG). Learning to engineer context is crucial for building robust, efficient, and reliable AI agent systems by managing the attention budget and ensuring high-signal information delivery.

Comprehensive guide for configuring Bring Your Own Key (BYOK) for major AI providers, including OpenAI, Anthropic, Google Gemini, Azure, and custom endpoints. This allows users to bypass default quotas, manage costs directly with the source provider, and utilize a wide range of specialized models within the Cursor IDE. It covers setup steps, security best practices, and cost monitoring strategies.

Deploying Machine Learning Models automates serving trained models to production by analyzing requirements, generating API and infrastructure code, and orchestrating deployment workflows with validation, error handling, and monitoring baked in.

Orchestrates PyTorch-native TorchTitan 4D parallelism (FSDP2, TP, PP, CP) for large-scale LLM pretraining, scaling from 8 to 512+ GPUs with Float8, torch.compile, and distributed checkpointing to boot Llama 3.1, DeepSeek V3, or custom models.

A comprehensive guide for selecting and optimizing embedding models for advanced vector search and retrieval-augmented generation (RAG) applications. This guide covers model comparison, chunking strategies, dimension reduction techniques, and provides runnable Python templates for both OpenAI and local Sentence Transformer implementations.

Feature Engineering Toolkit automates creating, selecting, and transforming dataset features to improve machine learning models; use it whenever you need new interactions, automatic selection, scaling, or encoding to boost accuracy and interpretability.

A comprehensive toolkit for evaluating the performance of code generation models. It supports benchmarking against industry standards like HumanEval, MBPP, and MultiPL-E across multiple languages. Use this tool to quantitatively compare different LLM coding abilities and measure code synthesis quality using pass@k metrics.

Evaluates large language models across 60+ academic benchmarks (MMLU, HumanEval, GSM8K, TruthfulQA, HellaSwag) with standardized prompts and metrics, making it easy to benchmark HuggingFace or API models, compare releases, and track training progress.