llm-prompt-optimizer

What problem does it solve?

Prompts for LLMs are easy to write but hard to make reliable, repeatable, and efficient. This skill provides a structured approach to transform vague prompts into precise instructions, reducing hallucinations and token waste while improving consistency across models.

Core Features & Use Cases

• RSCIT-based prompt design: define role, situation, constraints, instructions, and templates to control model behavior.
• Chain-of-Thought (CoT) and few-shot patterns: enable stepwise reasoning and reliable patterns across tasks.
• Structured output guidance: enforce explicit output formats (JSON, bullet lists, tables) for predictable results.
• Use Case: craft a system prompt for an AI agent that must produce a JSON schema and a short rationale for each decision.

Quick Start

Provide the prompt you want to optimize and specify the target model to begin the optimization.