Prompt Engineering Guide

Adoption assessment: Recommended as an introductory map and reference manual for prompt engineering, but it leans more toward literature review and conceptual explanation rather than a plug-and-play code library. What it actually does: The guide systematically organizes dozens of techniques such as zero-shot, few-shot, chain-of-thought, tree-of-thoughts, meta-prompting, and multimodal prompting, along with model settings and examples for tasks like coding, reasoning, and math. It helps users understand “how to ask” for better model outputs rather than providing a deployable tool. Barriers and cost: Completely free and open-source content with no API calls or paywall. Reading requires basic NLP literacy; best suited for developers, researchers, and product managers seeking systematic understanding. Not ideal for those who only want ready-made prompt templates without caring about the underlying principles. Social consensus and discussion quality: Frequently shared and bookmarked on X and Zhihu with high view counts and likes, but deep testing or comparative reviews are rare.