How does it handle structured outputs like JSON?

It provides a dedicated Structured Output Design workflow that includes schema definition, format enforcement prompts, and output validation steps.

How does this skill help reduce LLM API costs?

The Prompt Optimizer tool identifies redundant context and token waste, suggesting concise alternatives that maintain performance while lowering costs.

Can I use this to improve my RAG pipeline?

Yes, the RAG Evaluator measures metrics like context relevance, retrieval precision, and answer groundedness to help you tune your chunking and retrieval strategy.

Does it support agentic patterns like ReAct?

Absolutely. The Agent Orchestrator parses agent definitions, validates tool configurations, and provides ASCII or Mermaid visualizations of execution flows.

Senior Prompt Engineer

Name: Senior Prompt Engineer
Author: zhangzhang-111-i

byzhangzhang-111-i

0•

データサイエンスとML

Optimizes LLM prompts, evaluates RAG systems, and designs complex agentic workflows with standardized engineering patterns.

The Senior Prompt Engineer skill equips Claude with advanced tools for architecting high-performance AI systems. It provides automated prompt analysis for token efficiency, a specialized RAG evaluator to measure retrieval quality, and an agent orchestrator for validating complex ReAct or Plan-Execute patterns. By integrating established prompt engineering workflows—such as structured output design and few-shot example curation—this skill helps developers build more reliable, cost-effective, and accurate LLM applications.

主な機能

01Structured output design with JSON schema enforcement patterns

020 GitHub stars

03Agentic workflow visualization and tool configuration validation

04RAG evaluation metrics for context relevance and answer faithfulness

05Comprehensive library of few-shot and reasoning prompt templates

06Automated prompt analysis for token efficiency and clarity scoring

ユースケース

01Optimizing production prompts to reduce token costs and improve response consistency

02Architecting and debugging multi-step autonomous agents and tool-use workflows

03Benchmarking and improving the accuracy of Retrieval-Augmented Generation pipelines

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add zhangzhang-111-i/claude-skills111 senior-prompt-engineer

For use in Claude.ai and ChatGPT

主な機能

01Structured output design with JSON schema enforcement patterns

020 GitHub stars

03Agentic workflow visualization and tool configuration validation

04RAG evaluation metrics for context relevance and answer faithfulness

05Comprehensive library of few-shot and reasoning prompt templates

06Automated prompt analysis for token efficiency and clarity scoring

ユースケース

01Optimizing production prompts to reduce token costs and improve response consistency

02Architecting and debugging multi-step autonomous agents and tool-use workflows

03Benchmarking and improving the accuracy of Retrieval-Augmented Generation pipelines

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add zhangzhang-111-i/claude-skills111 senior-prompt-engineer

For use in Claude.ai and ChatGPT