Openai Prompt Engineer

This is essentially a reference guide packed into a skill. You get templates and patterns for chain-of-thought, few-shot prompting, and structured output formats, plus model-specific tweaks for GPT-5 (reasoning effort, agentic control) and Claude (XML tags, progressive disclosure). It's most useful when you're stuck iterating on a prompt or need to teach someone the difference between zero-shot and few-shot without linking them to a blog post. The examples are concrete enough to copy and adapt. Honestly, the value here depends on whether you want this baked into your Claude context or if you'd rather just bookmark a prompting guide. It's comprehensive but takes up context budget every time you invoke it.

Install to Claude Code

npx -y skills add jamesrochabrun/skills --skill openai-prompt-engineer --agent claude-code