when-debugging-ml-training-use-ml-training-debugger

Prompt Architect is a framework for developers to systematically analyze, refine, and optimize prompts using evidence-based techniques. It helps improve AI response quality and consistency by identifying anti-patterns and validating changes through A/B testing. Use it when you need to refactor an underperforming prompt or design a new, effective one from scratch.

This skill provides expert guidance for distributed training using Microsoft's DeepSpeed library. It helps developers implement optimization techniques like ZeRO stages, pipeline parallelism, and mixed-precision training. Use this skill when working with DeepSpeed features, debugging code, or learning best practices for large-scale model training.

This skill provides comprehensive performance profiling to measure, analyze, and optimize application performance across CPU, memory, I/O, and network dimensions. It helps developers identify bottlenecks, perform root cause analysis, and implement optimizations using tools like perf, Instruments, and clinic.js. Use it when you need systematic performance improvement through baseline measurement, detection, and optimization phases.

This skill provides comprehensive performance analysis and bottleneck detection for Claude Flow swarms, helping developers identify optimization opportunities. It offers real-time monitoring, profiling of swarm operations, and generates detailed reports with actionable recommendations. Use this skill when you need to diagnose performance issues and improve the efficiency of your Claude Code applications.