prompt-engineering

Core prompt-engineering protocols mandatory for ALL prompt-related tasks. Provides LLM prompt engineering best practices: clear/direct, multishot, chain-of-thought, XML tags, prompt chaining, long context, extended thinking. Includes guard rails: reduce hallucinations, increase consistency, mitigate jailbreaks, handle streaming refusals, reduce prompt leak, keep in character. Scope: general prompt engineering principles and best practices. Triggers: all prompt-related tasks.

// tip: Run this command in your terminal to install the skill