prompt-rewriter

The Prompt Rewriter is a professional-grade tool designed to transform vague or ineffective prompts into high-performance, structured instructions for AI models. It acts as an expert prompt engineering consultant, helping users bridge the gap between their initial intent and the AI's actual output. By evaluating inputs against core quality dimensions—Clarity, Specificity, Structure, Completeness, and Usability—the skill provides a comprehensive diagnostic report and clear, actionable rewrite options that consistently produce higher-quality results.

• Multi-dimensional analysis: Scores prompts from 0-10 based on structure, clarity, and goal alignment to identify exact weaknesses.

• Strategy-based variations: Generates three distinct versions of your prompt: a Conservative Fix for maintaining tone, a Technique-Enhanced version for complex tasks using patterns like Chain of Thought (CoT) and Few-Shot learning, and a Simplified version for accessibility.

• Expert pattern application: Automatically incorporates industry-standard prompt engineering frameworks such as Role-Based prompting, output template specification, constraint definitions, and step-by-step reasoning instructions.

• Educational feedback loop: Provides detailed explanations of why specific changes improve model performance, helping users internalize prompt engineering principles over time.

• A/B testing support: Guidance on how to iterate through different prompt approaches to systematically find the most effective version for specific use cases.

• Ideal for power users, prompt engineers, and developers looking to standardize AI interaction quality.

• Inputs: Vague natural language requests, rough drafts, or failing prompts that require structural or context-based improvements.

• Outputs: A diagnostic score, multiple refined prompt variations, clear explanations of applied techniques, and recommendations for optimal deployment.

• Constraints: Focuses on prompt structure and logic; ensures that rewritten prompts remain within token efficiency bounds while maximizing accuracy and output formatting.