prompt-engineering

Design LLM prompts for reliable structured output generation by applying system prompts, few-shot examples, and chain-of-thought instructions to enforce predictable and secure model responses.

Build a RAG pipeline with robust chunking and reranking strategies to efficiently manage context windows and improve retrieval accuracy for your AI features.

Tool calling for agentic workflows implements function calling to connect LLM APIs with external tools, enabling models to autonomously execute predefined functions and complex multi-step tasks.

Optimize token usage and model routing for cost-efficient LLM features by applying efficient context window management and model selection strategies across OpenAI, Anthropic, and local models.

Defend against prompt injection vulnerabilities in AI applications by applying security patterns and best practices that protect system prompts and ensure safe LLM integration.

Yes, you can use prompt engineering patterns with local models and OpenAI APIs, as the skill provides integration best practices for multiple LLM providers to ensure broad compatibility.