prompt-engineering-guide

The Prompt Engineering Guide is a curated collection of guides, papers, lessons, notebooks, and resources focused on prompt engineering, context engineering, Retrieval Augmented Generation (RAG), and AI agents. It serves as an educational reference for researchers and developers looking to better understand and optimize interactions with large language models (LLMs). The guide covers foundational topics such as zero-shot and few-shot prompting, chain-of-thought reasoning, self-consistency, prompt chaining, tree of thoughts, and retrieval-augmented generation. It also explores advanced techniques like automatic prompt engineering, active-prompt, directional stimulus prompting, program-aided language models, ReAct prompting, multimodal CoT, and graph-based prompting. Application areas include function calling, code generation, data generation, synthetic dataset creation for RAG, classification, information extraction, mathematical reasoning, text summarization, question answering, and adversarial prompting. The guide provides model-specific guidance for ChatGPT, GPT-4, Gemini, LLaMA, Code Llama, Mistral, Mixtral, OLMo, Phi-2, and Flan. Additional resources include a curated paper library with overviews, approaches, and collections; a prompt hub organized by task type; tool recommendations; Jupyter notebooks; datasets; and risk/misuse considerations such as adversarial prompting, factuality, and bias. The guide is available online at promptingguide.ai and can also be run locally using Node.js and pnpm. It is actively maintained and has been featured in publications such as the Wall Street Journal and Forbes. The project is licensed under MIT.