What problem does it solve? Prompts often waste tokens, produce inconsistent outputs, or lack clear structure, and RAG or agent systems are hard to validate before deployment. This Skill provides static analysis, evaluation metrics, and design patterns to systematically improve prompt quality and agent reliability. ## Core Features & Use Cases - Prompt Optimization: Analyze prompts for token count, cost estimation, clarity scores, ambiguity, and redundancy, then generate optimized versions with concrete suggestions. - RAG Evaluation: Measure context relevance, retrieval precision, answer faithfulness, and groundedness against evaluation question sets. - Agent Validation & Visualization: Validate agent YAML/JSON configurations, detect potential infinite loops, estimate token costs per run, and render workflows as ASCII or Mermaid diagrams. - Use Case: Before shipping a customer-support agent, run the orchestrator to validate tool configurations and estimate monthly token costs, then use the prompt optimizer to cut redundant instructions and reduce per-query spend. ## Quick Start Ask the assistant to analyze your prompt file for token usage, clarity issues, and optimization suggestions using the prompt optimizer script.
