Prompt Engineering

소개

Elevate the effectiveness of agent interactions and sub-agent tasking with a comprehensive toolkit of advanced prompt engineering techniques. This skill provides structural frameworks for few-shot learning, chain-of-thought reasoning, and systematic prompt optimization to ensure production-grade outputs. By incorporating psychological persuasion principles—such as authority and commitment—and Claude-specific best practices like token-efficient context management, it enables the creation of robust, controllable instructions tailored for complex agentic workflows.

주요 기능

• Context window management and token efficiency strategies
• 79 GitHub stars
• Progressive disclosure and instruction hierarchy patterns
• Few-shot and chain-of-thought reasoning frameworks
• Systematic prompt optimization and template systems
• Application of psychological persuasion principles for agent compliance

사용 사례

• Designing robust production prompts for sub-agents and automated LLM interactions
• Optimizing existing prompts to improve accuracy and reduce token consumption
• Creating high-compliance agent commands using established psychological principles