Prompt Engineer

When your chatbot gives inconsistent answers or your AI agent hallucinates information, this is what you reach for. It applies the CLEAR framework (Context, Limits, Examples, Action, Review) and specific techniques like chain-of-thought prompting and constraint engineering to turn vague instructions into precise system prompts. The skill claims 40-60% improvement in response consistency, which tracks with real-world experience when you add concrete examples and edge case handling. Best for designing agent instructions and chatbot system prompts. Skip it if you need automated optimization loops, that's what automatic-stateful-prompt-improver does. The anti-patterns section alone (prompt stuffing, over-constraining) is worth reading before you write your next system prompt.

Install to Claude Code

npx -y skills add erichowens/some_claude_skills --skill prompt-engineer --agent claude-code