Ai development

Learn how to systematically improve your AI prompts through a five-step evaluation workflow. This guide covers drafting prompts, creating eval datasets, generating responses, grading outputs, and iterating for better performance—transforming prompt engineering from guesswork into data-driven optimization.

Learn how to use Claude's built-in text editor tool for file manipulation. Understand schema versions, implementation requirements, and practical applications for file editing.

Learn how to use XML tags to add structure and clarity to Claude prompts. Improve prompt parsing with clear boundaries for code, documentation, and data.

Learn how to improve Claude's output quality by being specific in your prompts. Master output quality guidelines and process steps to get consistent, high-quality results.

Learn how to use hooks in Claude Code to address common AI development weaknesses. Implement TypeScript type checking and query duplication prevention with practical examples and code.

Master the art of writing clear and direct prompts that get better AI results. Learn why the first line matters most, how to use action verbs effectively, avoid common mistakes like vague questions, and see real examples showing 69% improvement from this single technique.

Learn how to distribute skills across teams and organizations using repository commits, plugins, and enterprise managed settings. Discover how to configure custom subagents to use skills.

When skills don't work as expected, the problem usually falls into a few predictable categories. Learn systematic troubleshooting approaches for skills that don't trigger, load, or execute properly.

Claude Code offers multiple customization options. Learn when to use skills versus CLAUDE.md, subagents, hooks, and MCP servers, and how they complement each other in a typical development setup.

Master advanced skill techniques including metadata fields, tool restrictions, and progressive disclosure. Learn how to organize complex skills across multiple files while keeping context efficient.